Okay

>> Okay.

so let's begin.

it's a little late.

actually today's presentation begins, what is a series inadvertently of the interface of

genomics and medicine.

so next week we'll be both sort of clinical and genomic aspects of neurologic disease,

particularly dementias, Alzheimer's, Parkinson's, schizophrenia a little bit.

and then the following week Francis Collins and -- what's her name.

any rate, Sharon MILGRAM are going to tuck about cystic fibrosis and that's also from

the standpoint of the profound interface of genomics and human disease.

so I don't know, when I think about genomics, I think of something like this, which at one

point when someone gave me this slide they said it was for real.

it was a allegedly some engineer who worked on an oil platform or on a ship or

something and saw the hole of an iceberg, turn out it's a fake.

a reassembly.

's the idea that what we see almost in medicine historically is invariably the very tip of

the iceberg and then slowly work our way down to where the real underlying substance

is.

and is with this it's under the water.

and so the questions that come to my mind and just by way of a very brief background for

those of you who may not know this, so it seems where we are now is people ask what

are good genes and what are bad genes.

who has which.

and what does it mean?

I guess that's the big question.

what does it mean if you have susceptibility genes?

Should you take it all that seriously or not?

And what can be done about it?

The history of this development is condensed into three lines here.

one is over a hundred years ago where people who had what now we know as mendelian

inherited diseases who kind of look different, sometimes very dramatically so in terms of

their bone structure, their brains, their organs.

and if they weren't lethal they were often considered medical curiosities.

and they appeared in circuses and side shows and they were the object of great prejudice,

COZIMOTO was probably a good example of something like that.

about a hundred years ago sir ARCHIBALD Garret wrote a classical paper which I'm

going put on the website for those interested tomorrow where he came one the idea that

no two individuals are exactly alike chemically any more than they are structurally.

that came from studies of rare, rare diseases, particularly one I'll call captainURIA.

he created this label of in borne errors of metabolism.

he knew these were inherited but of course in those days that's pretty much where it

ended because nobody knew from genes or anything like that.

then for the next hundred years up until only about 15 years ago we persisted in this idea

of identifying inherited diseases by virtue of a gene product, not by virtue of a gene.

then comes the genome which now we're right at the edge of where the water and the

iceberg exist, the whole world lies underneath it, nobody knows how deep it is, how wide

it is, and what all of its ramifications are.

but it certainly is now and facetiously but true, forever because there is no way back.

so now we think in terms of the genome and inherited and acquired diseases,

susceptibility.

so how do you really begin to translate this sort of thing from a big conceptual picture

into reality?

That's what pretty much is the topic of today's presentation.

so we're very fortunate to have two speakers who are ready and willing to answer your

questions and hope that you keep them in mind.

they're going to do like a tag team, speak at different intervals.

so the first speaker is Leslie Biesecker who is chief and senior investigator of the genetic

disease research branch of the genome institute.

He's a physician who was trained in pediatrics, in genetics, came here in the 1990s,

received a great deal of attention by virtue of his work on some of the diseases that you'll

hear about today.

and he also participated in developing a brochure for DNA testing that was handed to

survivors and families from the world trade disaster, which I mention because it's that

kind of information transfer which is really critical to having the public understand what

all of this iceberg event is about.

and so our speakers Julie Sapp who is a genetics -- what are are you officially, genome

counselor.



>> Genetic.



>> Genetic counselor and she deals to a great extwenty the families of people who have

mendelian genetic diseases and we'll hear much more about that.

so there's going to be a considerable range from technology and application to clinical

diseases to families and to real life.

so who is up first?

Thank you.



>> Thank you all for coming.

is this microphone on?

Can you hear in the back?

Okay.



>> Can you hear all right back there?



>> No.



>> No.

all right.

is that better?

Can you hear in the back?

>> You can always move up to the front.



>> I can use the -- I can use this one.

is that better?

Does that one work better?

Doesn't seem to have as much volume.



>> No problem.

that's okay.

can you give me a little more volume on this?

Yeah.

how is that?

Can you hear in the back now?

Okay.

all right.

let's get going.

so Julie and I want to do as he said a little tag team approach to talk with you about how

we approach questions of disease and gene identification as well as care of families

affected by these disorders.

What we hope to convince you of is the complimentary roles that a geneticist and clown

has in this translational research environment of a genetics research program.

please interrupt as often as you can.

I won't speak for Julie but I definitely get tired of hearing myself talk so if you guys have

questions yell or wave your hand or something and we'll be happy to stop.

so two themes today of what I want to talk to you about are the notions of what is

translational genetics and what is translational genomics?

And we try to practice both in our research endeavors.

first thing I want to do is go through some examples of how we use a translational model

and what we call the syndrome family approach to disecting the pathophysiology of

malformations and other genetic disorders in humans and the relationship of how gene

variants relate sometimes clearly, sometimes not so clearly to the phenotypes that they

cause.

And segue from that into translational genomics, scaling genetics and thinking about

translational genomics which is an entirely new enterprise.

so we want to provoke y'all to think about what you can and can't do in the clinic.

one thing I go crazy with is when anybody suggests anything like we've always done it

that way.

what we're about here at the NIH should be about always trying to turn the apple carts

upside down and trying to do things new ways to answer important questions.

so our job is incredibly simple.

all we have to do is answer these four questions.

these are the questions that usually parents of children ask when they bring a kid into the

clinic who have some problem.

what is it that's wrong with the child?

Julie is going to talk extensively about this notion of what a diagnosis is and what it

means to a family.

what caused it.

usually a lot to that question.

it's only three words but usually there are many facets of that question when parts ask us

that and we can address some of those facets but not others.

What is going to happen in the future T prognosis, the natural history of the disorder that

the child has underlying the importance of diagnosis and correctly categorizing patients.

Of course we want to know what we can do about it.

Are there any interventions that will modify or ameliorate the phenotype that they or their

children have.

so what we do as translational researchers is in the end kind of shameless.

that we exploit existing techniques and technologies to answer these four questions that

families ask, that's all there is to it.

We are not the people who invent the ZIPITY doo DAH machines I'll show you later in

the talk.

we don't invent new technology paradigms.

We take the technologies the basic scientists develop to understand what the potential

applications are of those technologies and then translate those into the clinic so that we

can answer those questions that the families ask.

so examples.

this was lady who called me about 12, 13 years ago at the NIH because I had published

one little case report on a rare phenotype called PALLISTER hall syndrome.

she found it through the literature and she said my doctor says I have this.

I asked her what the manifestations were and she explained that to me and I asked her

does anyone else in your family have this?

She said oh, yes, there's 20 or more of us who have this, which at that time was stunning

because at that time there was only one or two families in the literature with one or two

people each who have this phenotype.

that gives us the tool to approach this problem genetically and dissect the molecular basis

of that phenotype.

a different kind of problem is the second one, a much more common problem.

this is a patient who presents into a clinic and says that I have high cholesterol.

this is a rampant problem in our culture.

this lady has a little bit more severe flavor than usual of it.

in fact, she's had several relatives who died from early onset heart disease.

she has a grandson who is only four who has already been diagnosed with high

cholesterol so here you're getting a flavor of something a little atypical about this family.





[off mic]



>> We don't know what the value is.

grandma knows she has high cholesterol but we don't know the number but we'll know in

a few weeks because she's coming up to see us.

but 4 is essentially never seen at the age of four years old.

extremely rare.

she's successfully treated with a number of drugs known to reduce cholesterol levels and

prior to our evaluation of her and her family was known not to have anything other than

garden variety high cholesterol.

so how do we do this?

Take these families who have these problems and we need to translate that into molecular

genetics answers.

The paradigm sort of looks like this, this is the way we've always done it.

this is what we do.

we delineate the phenotypes of the disorders.

it's very important to properly lump and split phenotypes into categories in order to do the

downstream research.

we carefully phenotype the patients and that's what the clinical research center is all about

doing, something that we can do very well here, a lot better than a lot of other places in

the world.

now, when we take samples from the patients, generate some flavor of genetic data,

determine causation, no small question sometimes.

Then we take those results because again we're translational researchers.

We take those data, those conclusions and we go back into the clinic and return those to

the family so the family gets something out of participating in this research and we do

this in iterative fashion to continuously improve our understanding of these disorders and

help the families.

Then we take whatever we've learned and what always is the case is we learn something

and it raises one, two or three more questions.

so then we reformulate those new questions into research questions that we can attack or

exploit and we design follow-on studies.

We sort of as you would say turn the crank in iterative fashion to answer the family's

questions in this way.

what I would suggest to you is this is not the only way.

just because this is the way it's been done essentially since Garrett laid out this paradigm

a hundred years ago there's no reason it has to be done this way, I would suggest some

other ways we can do it that might be even better.

In human translational research we have some problems that the basic scientists don't

have.

and those problems include number one, we cannot manipulate the genetics of our

system.

we don't set up our crosses, right?

Mouse geneticists get to do that.

the genetics is often far from simple and it can be difficult to tease out the inheritance

patterns of phenotypes in human families.

We definitely have limited access to tissues and samples and different time points.

we don't always have early developmental samples.

we don't always have that brain tissue that you would like to have to do some functional

assay of some gene variant.

We have to work with what we have.

it's certainly expensive and it is definitely slow because of the generation time in humans

and intrinsic limitation as well as the regulatory burden of doing human subjects research.

but there are opportunities that no animal system can match.

the detail -- the level of detail of human phenotyping with the exception of the cellular

mapping and sea elegance worm, there's no organism that is as well studies as is the

human.

so we have a huge advantage there.

the phenotype ranges are enormous from extensively severe to extremely subtle

phenotypes that you can't model in an animal because they are so subtle.

it's something you can do quite readily in humans.

and there's -- in humans. There's a gigantic database of knowledge.

we can go into the literature and look back 25, 50, 100, 200 years and find really good

data on the phenotypes we're interested in and learn from the people who studied this

before us.

so how do we do translational genetics?

We hypothesize that patients have overlapping malformation syndromes.

that is, malformation syndromes that include manifestations, component malformations

and that these different syndromes that the patients have partially overlap.

we suggest that the reason why these syndromes overlap is because these disorders are

caused by mutations in genes that function in genetic developmental pathways, in

coupled pathways.

so that's why the patients have overlapping phenotypes because the different disorders are

caused by mutations in genes that are perturbing the same pathway.

so how does this work in practice?

Okay.

so the lady -- patient one who contacted me PALLISTAR hall syndrome?

Three component malformations.

The first is a central nervous system malformation on this MrI image of a -- MRI image

of a brain is a massive tumor not supposed to be there.

s that an overgrowth of tissue in the center of the brain that can cause endocrine problems

or seizures.

Most of the time though it is asymptomatic.

it can have a peculiar form of polydactyly extra digits.

he's up to seven so he has hepta dactylY here.

we have seen patients with as high as 8 digs and then airway manifestations, the most

common is a cleft in the epiglottis.

so those make the syndrome.

the puzzle about this disorder is it has a phenomenal range of severity.

so we have a number of patients that come from simplex families with just a single

affected case where the patient has a congenital lethal presentation, patient is born, the

anomalies are obvious and within hours or day or two the patient is deceased.

Then we have other patients like patient one who called me, who was in her 60s and this

is a relative of hers who is -- you would have a hard time convincing a lay person that

these two people have the same disorder, right?

Because of incredible difference in severity.

yes, they do.

so how do we do this?

Well, the short of it, this is a compression of about oh, I would say about 1 or 2 person

years worth of work is this slide, the paradigm is called positional cloning.

what you do is phenotype and sample a pedigree of affected family members, you collect

blood from them.

you lyse late DNA and -- isolate lots of DNA markers.

you take the alleles of those markers and you look in the pedigree to see which alleles of

which markers co-segregate with the phenotype.

lots of statistics and I won't put up formulas because people always go to sleep when you

do that.

but suffice it to say mathematically you can quantitate the likely hoods of these

probabilities and that tells you where the gene probably lies.

and you refine that mapping through a number of techniques and then you start to

sequence candidate genes to look for mutations in the genes that may cause the

phenotype.

so turned out for this phenotype -- wrong button.

that they -- the patients had a mutation in the gene called GLI-3.

And essentially all patients who have this -- who have this triad of the three have a

particular kind of a mutation in this gene.

so that gives us step one of our translational algorithm.

so we have take an group of malformations, we have recognized them as being in a

syndrome and we found the gene alteration that causes them.

problem was was that this gene was already known to be mutated in patients who have a

different disorder, a disorder that's distinct from palLISTER hall syndrome.

that's the Greg cephal polysyndactyly syndrome.

This is not GREIG the composer.

two guys.

This disorder has different manifestations.

Wide spaced eyes, a different polydactylY usually with severe cutaneous webbing

between the digits and duplicated great toes and it is quite distinct and they don't have a

epiglottis.

so what explains this discrepancy because these two phenotypes are very different from

each other.

the other thing is that greig CEPHALOPOLYSYNDACTYLI syndrome, which is GCS

because that's a mouthful has its own range of severity.

when we have mutations in a gene you like to think that's a range of mutations ranging

from mild to severe ones perhaps all the way through to dominant negative mutations and

that that correlates with the range of severity of the phenotype.

I sort of call this the usual suspect, if you know the movie.

we like to think these will map nicely in approximate linear fashion and make perfect

sense.

in these disorders they don't because we have mutations in the genes and patient whose

have two qualitatively distinct phenotypes.

so what's going on here?

You start to look at more patients and do what we call genotype, phenotype correlation,

as certain large numbers of patients, accurately phenotype them and try and understand

the relationship between the two.

what do you see when you do that?

Turns out that patients with PALLISTER hall syndrome almost always have one class of

mutations in this gene.

whereas the patients with GCPS have all different kinds of mutations.

There are end frame deletions, translocation, large deletions, medium size deletions,

missense mutation, slicing mutations and premature truncation mutation.

All of them in GCPS and essentially only one in palLISTER hall.

what is the deal?

Around the time we made this discovery turns out a group in San Francisco had identified

the fly homolog of this gene and determined that it was a gene that functioned in a

genetic pathway downstream of a gene called hedge hog patched in smoothen and that

there was a huge range of severity of this phenotype in the fly -- phenotype in the fly.

what they also showed was that this protein in the fly called CI, cubicTUS interruptUS

was a transcription factor downstream of hedge hog signalling such that the protein was

cleaved into a shortened form when hedge hog was absent and transferred to the nucleus

in a longer form when it was -- when hedge Hodge was present and this form of a gene

activates downstream targets and this form of the gene represses them.

so what's going on in these phenotypes?

Look what we have got here.

if you take all of these mutations in this gene, the overlapping class that both disorders

have, they don't distribute randomly across the gene.

almost all of the patients with GCPS have truncation mutations in the five prime or

median terminal end of the protein and when you get just past this domain of the protein

which is the DNA binding domain, the ZINC finger domain of the transcription factor,

the Phenotype flips the palLISTER hall syndrome with one exception, stays as

palLISTER hall syndrome and when you get to this end of the protein it flips back to

GCPS.

this point in the human protein is exactly analogous to this point in the fly protein.

so what's going on is that the human mutations are mimicking the bifunctional aspect of

this protein as it is processed in the fruit fly explaining why the two different mutations

have different consequences.

so now we can sort of start to expand our algorithm.

we have two syndromes here that have partially overlapping manifestations but are

distinct.

both caused by mutations in the one gene and processing of the protein product of that

gene and the differences in the way that is processed explains the relationship of the two

phenotypes.

except that again, there is a spectrum of severity of this GCPS phenotype.

what's going on there?

We found a subset of patients, 5 to 10% have this phenotype also have mental retardation

and seizures.

not explained by the phenotype.

so we took another tool of our basic science colleagues called a ray CGH.

we applied that tool to this problem because we hypothesized that deletions of this gene

that are much bigger than the gene itself cause the phenotype of GCPS but also cause

other troubles.

Those other troubles are mental retardation and seizures.

so this is an array CGH result, one of the patients, and the probes are outlined along the

genes, the chromosomes here and here is chromosome 7, this is a custom array so you

can see there's hundreds and hundreds of probes here but the density goes up to

thousands, right in this area of chromosome 7 which is where the gene lies.

If you zoom in on that you can see normal chromosome dosage coming from the P arm

here and then a big deletion, probably about a mega base in size.

Here is GLI-3 here and flips back to normal.

s that patient that has about a mega base, a million base pairs of DNA deleted much

bigger than the gene itself.

so Julie will talk more about this later but we named this phenotype a new designation,

we call this the Greg contiguous gene syndrome.

you have to pity the parents who have to have all those syllables attached to their child

but it is distinct from garden variety GCPS, caused by ha flow insufficiency of GLI-3

plus haplo deficiency of neighbors.

so extending the model, going on to another overlapping syndrome is the Mick HUSEN

syndrome.

it shares but also has in addition congenital heart defects as a genital urinary

malformation in females.

turns out this phenotype is almost unique to a group of people who live just north of here,

the old order Amish of Lancaster county.

this is inherited as a recesssive phenotype described 40 years ago by Victor MEKUSI.

autosomal pattern it is often lethal because of the heart defect.

so again we did our positional cloning drill, identified the gene and it was in a novel gene

which we named MKKS and we identified missense mutations in that gene.

so now we can extend our algorithm and have MKAS here.

so what we needed to but tress this argument was access to additional patients buzz when

you do -- because when you do gene mapping in an inbred kindred you have only one

allele of a gene.

Because inbred, one mutation in that cohort, you need additional mutations so we started

to follow up on literature reports again because all that information that's out there for us

to use.

of non-amish MKKS patients.

We contacted colleagues about patients we got three answers either they were lost to

follow-up which happen as lot.

they were deceased because the disorder is severe, the rest that they have did have

information on they told minimum my patient's diagnosis isn't MKS any more.

they changed the diagnosis from MKS to another disorder which is odd to have happen.

what happened?

Over time MK was diagnosed when they were newborns but through the school years

they became obese, it was clear they had cognitive problems some had mental

retardation, developed diabetes and pigmentary retinopathy.

so their diagnosis was changed to a syndrome called BARTA beetle syndrome.

another pleiotropic syndrome that contains genital anomalies and heart disease and also

includes the time or aged dependent manifestations the patients developed.

Average age of this diagnosis -- this disorder is nine years.

The question is, okay, now, look what's happened to us.

We have added another syndrome with partially overlapping phenotype with additional

manifestations.

The question is, is this allelic to MKS?

Turns out it is.

but mutations in the MKS gene, the gene that causes this other disorder in Amish, in the

non-amish causes BartA beetle syndrome.

MKKS is a family of genes that cause barA beetle syndrome and these genes function in

a biological pathway called the primary sill yum.

basic cellular structure.

and it's subsequently turned out that cilia, the basal body of the sill yum is necessary for

the mechanical processing of the GLI-3 transcription factor.

so that's this processing is pathophysiologically coupled through the sill yum to the gene

products that cause Barb barnA beetle syndrome and the MKS validating this approach.

so here is the lessons that we have learned from this work, probably took about ten years

to this phenotypes and genotyping.

we accurately delineated a phenotype Hal palLISTER hall syndrome.

and developed a prognostic and diagnostic tool for all these disorders.

delineated the cause of MKS, the first cause of barnA beetle syndrome and have

developed diagnostic and management algorithms now in textbooks that clinics can use

to manage their patients.

One consider very self-satisfied with this.

we have done a lot of approaches over the years.

clinical, molecular biology, genomics, to try to understand the manifestations of the

phenotypes in our patients.

One shouldn't be so self-satisfied.

Does anyone know what this is?





[off mic]



>> The ONFELOWS of Delphi.

this was a statuary in ancient Greece, it was the center of Delphi and they were a rather

self-satisfied bunch of people and they thought this was the center of the world.

from this artifact and this attitude that the DelphiANS had is to rise to use the phrase of

naval gazing.

someone was a little self-satisfied looking too inward we say they're naval gazing.

we don't want to be accused of that so how do we take what we have learned and apply it

to other problems outside?

What have we learned overlapping pleiotropic phenotypes that caused by genetic

pathways.

genes pathways don't connect one to one.

when I started training we thought if we just knew the genes we wouldn't have to or

worry about diagnosing patients because the gene mutations would tell us exactly what

everyone else.

complete completely wrong.

more complex than that.

the other thick we learned is there is a continuous spectrum between monogenic disorders

and oligogenetic disorders and multi-factorial disorders, not cleanly delineated into

simple and complex genetic diseases.

don't allow people to pigeon hole biology.

human beings love to dichotomize things.

We like to thee there are two kinds of Deess in humans.

Rare diseases caused by rare variants or mutations in genes.

high penetrants, then other diseases, common diseases caused by common variants and

those disrders are caused by alterations in a number of genes that additively cause

patients to have a particular type of phenotype.

that's developed into these models of susceptibility.

There's alternative ways to think about this.

some think that susceptibility to common diseases is conferred primarily by alleles

common in the population patients have several of them and each allele contributes

additively to the phenotypes T multi-genic model.

Others suggested common diseases are the result of multiple rare alleles not shared

among patients an each independently have large phenotypic effects.

Truth is probably that it's somewhat in between.

there is a spectrum of gene effects from high penetrants to low penetrants from common

to rare.

and then different phenotypes contribute additively to aggregating a rat phenotype we see

in the clinic.

and I like to use the sort of proportionate I didn't know and yang model if that there are

some rare -- some phenotypes which rare varyians contribute most the not all variants of

the disorder and common variants don't have much to say about it and there's probably

some that have a complete spectrum of this effect.

the phenotype is heterogenous and is an add mixture of abnormalities caused by a few

rare variants and some common variants and different add mixtures and different

patients.

how do you approach the problem?

This kind of -- approach this problem?

This problem where you have a spectrum of effects and relationships of genotype to

phenotype subpoena a very difficult one to attack by any linkage or association approach

because it is so heterogenous.

other than the direct root force approach of sequencing to just go in to an individual

sample and interrogate large amounts of sequence.

how do you the do that?

The NIH supports the genome institute a number of NIH funded sequencing centers.

There are five of them currently.

they currently have the capacity to produce two times 10 to the 1 is 1st base pairs of

sequence per year.

I want you just to think about that for a second.

remember the entire human genome project, the goal of that project was to generate three

times ten to the 9th base pairs of DNA sequence in ten years.

three times ten to the 9th in ten years and the senters now can produce 20 times 10 o -- 2

times 10 to the 11th per year, that's equivalents to 60 to 70 genome mammalian genomes

annually.

that's not enough.

what's happening now is technological developments are occurring such that this is going

to scale by several more orders of magnitude.

There's several companies that have developed originally with NIH starter grants, various

flavors of what I like to call solid phase sequencing which is massively parallel

microsequencing reacts that occur on chips, that generate millions and millions of short

sequence reads in C-2 on a chip that are red by lasers and computers.

these technologies in a single run of a single sequencing machine can generate about 10

to the 9th base Paris of DNA sequence and the run cost is about $100,000.

Sounds like a lot of money, it is but again remember the genome project was 10 years,

three times ten to the 9th base Paris and about a billion dollars.

so the costs are falling exponentially.

so a confluence of advances are occurring.

we're developing a more sophisticated understanding of this complex genetic architecture

of human disease, we have the technological development coming down the pipe where

we can generate higher and higher amounts of sequence data every year.

how can we take advantage of that?

What we need to do is exploit these questions of geneticA texture and technological

advances and develop new approaches to clinical research.

large scale medical sequencing, use a clinical research tool.

question is how do we use it?

It has attractive ability to detect common and rare variants.

When you gene up type an individual, you're going out interrogating a sequence variant

you only get out of the machine what you ask it.

You ask if is there an A or C at this particular point in the genome.

sequencing tells you what's there regardless of what it is.

You don't have to know it's there to look for it so it's sort of a model free or hypothesis

generating machine as opposed to hypothesis answering machine.

so you don't have to have assumptions about what kind of sequence alterationious expect

to find.

you find them all.

again, the repeatedly allow us to generate gigA bases of DNA sequencing in case where it

used to take years.

To exploit this we develop ad protocol to try and take advantage of this situation and

answer questions and develop a pipeline for how we could do this.

so we set out tow pilot large scale medical sequencing as a clinical research tool to

answer practical questions.

so in the NIH clinical center now we -- this patrol is underway and we're piloting it on a

scale that's practical an models whole genome sequence acquisition.

so we expect within 3 to 5 years there will be patients coming from the NIH clinical

center and we'll have the ability to sequence their entire genome as part of a clinical

research project to answer important questions.

the aims of this project are listed here, we need to develop a lot of infrastructure to try

and move this process forward because none of us know how to do this kind of research.

We need the use large-scale medical sequencing data to address biomedical research

questions or appropriate questions and we need to understand how in the world are we

supposed to enteract with human subjects top whom we are doing this?

We know really well, we have really good paradigms worked out for how to interact with

a subject who comes to us with a given disorder and wants to know the cause of it.

that's a completely different issue than a subject coming to the research protocol and we

troll through their entire genome and we have to go back to them and answer their

questions of what we found.

completely dynamic.

this is a transdisciplinary project involved investigators in our institute, the genome

institute, the national heart,lung and blood institute for reasons you'll see in a second.

the NIH intramural sequencing center up in Rockville.

I would encourage anybody who is curious about high through



Put sequencing to get yourself a tour of this place, it's quite a remarkable facility after

twin brook avenue.

so what are we going to do to these patients?

We bring in a cohort of a thousand people and we will phenotype them for a common

disorder@row sclerosis dislipid deemia, myocardial infarction.

it's clear there are rare mendelian forms of athrow sclerosis so it's a good example of a

disorder with the genetic spectrum of attributes we're looking at.

We're going to as our pilot sequence 400 genes for athrow sclerosis in this cohort of a

thousand people.

follow up studies to validate these finds.

Interpret the variant.

when weapon identify variants that we think are high penetrants, rare alleles we will

validate those and return the results to the patients.

this is the only protocol that I'm aware of in the world right now where patients are

undergoing high through put sequencing and getting results back.

yes.





[off mic]

Great question.

how we pick those 400 genes.

some of them are ringers, some are genes already known to cause athrow sclerosis like

low density lipoprotein reaccept for.

PCS K-9 et cetera.

Others are genes in the pathway with known genes that cause atherosclerosis.

Other genes are genes that have been shown say in a mouse to cause dislipid deemia or

atherosclerosis but no causing relationship has been found.

so any and all reason they just get plowed into the study.

in a way it doesn't so much matter which genes we select because in the end these

patients are almost certainly going to undergo whole genome sequencing we're

consenting them for that.

so this is just a pilot to allow us to get our feet wet.

but we salted it enough to get positive control which is we know we have.





[off mic]

We have enlisted three individual whose have high penetrants gene variants that we have

already detected and one has been returned, two will be returned to the family shortly.





[off mic]



>> So the question is how many controls?

We are not treating this as a case control cohort.

this is a -- we are modeling athrow sclerosis as a continuous trait.

The study, we're calling it if you will so one-fourth of the patients will fall into each of

these phenotypic bins so we have a complete spectrum from people who are disease free

to people who are already had myocardial infarction, or have undergone stint placement

or coronary artery bypass surgery for a serious lesion so we'll have the whole spectrum.

we're going to do this as a quantitative trait association, not as a case control.

that's more reality, these are continuous traits.





[off mic]



>> The question is what kind of informatic analysis --



[off mic]



>> That is the $64,000 question.

it is much easier to generate these data than to analyze them.

that's the hard part.

what we're doing now is we're -- we're sort of doing what we know how to do.

so we know how to troll through these sequencing and find stop mutations and nonsense

mutations and deleterious splice mutations.

That's easy, that's at the stage we're at now.

we're detecting a lot, as you would predict lots of missense variants.

That gets harder to decide what's abnormal and what's normal because there's a full

spectrum of allelic variation in these phenotypes.

that's going to require probably replication studies to validate those.

in vitro studies to figure out the biology of those variants.

We're going to have to develop that paradigm.

the other thing we have to do which I think is really exciting is nobody has any idea of

how on earth we're going to display these data.

We're going to develop 8 million sequence reads on these patients bidirectional sequence

on a thousand patients.

How are you going to display that so an investigator, clinician or scientist can go in and

look at that time variants, understand the relationship of the vary I can't want to all the

phenotypic data we're collecting on the patients and begin to tease that out and just the

ability to browse these data is a huge bioinformatic challenge so we're going to

collaborate with data and visualization researchers to develop because we don't know

how to do it.





[off mic]



>> The question is what are we genotyping here?

At this stage we want absolutely pristine DNA.

so we collected 40 MLs of fresh blood to isolate DNA.

so this is being done on unadulterated DNA from a buffy coat preparation.

We're banking lymphoblast cell lines from the patients and those are made available to

other researchers.

We're banking serum on all the patients and we have the ability to do follow up studies

with the patients so they can come back and we can come back in different ways.

Make this sort of into a resource so not just we can use it.





[off mic]



>> Okay.

the clinical evaluation by NIH standards is fairly brief.

we get the patients in and out of the clinical center in just over half a day which is a

minor miracle by NIH standards and they have a semiautomated family history

ascertained, a form driven medical history, a brief examination, we had quite an argument

in the design state of the study as to whether or not we would put a stethoscope on these

patients' chests.

really interesting argument.

I was adamant that we had to do that because for heaven sake it's a study of heart disease.

my colleagues in heart lung and blood institute convinced me that was gibberish and that

we were collecting bate data than -- because we were doing echo cardiography, CT

scanning, and electro cardiography and so they made me give that up and lots of clinical

laboratory tests.

you asked the question here about sequence.

how do we detect the variance?

The problem of detecting hetero zygous DNA variants is a current major bioinformatic

challenge.

most of the genome sequencing is done out of library where is you're sequencing

essentially one allele.

sequencing cloned DNA.

we're sequencing hetero zygous DNA and most variants are present in only one of two

copies and the non-trivial to teach a computer to see hetero zygous variants reliably and

specifically so that's a challenge.

the other ones we have to detect are frame ship mutations so sequencing hetero zygous

DNA, training a computer to recognize a frame shift is no small matter because in cloned

DNA you don't have to deal with that.

in humans you do.

we have to decide what is a disease gene.

just a trivia question, how many disease genes are there in the human genome?

Anybody know?





[off mic]



>> Yes.





[off mic]



>> 2500 was one.

30,000, she says.

Any other estimates?

There aren't any.

we don't have disease genes.

We have genes that have normal physiologic functions that when mutated cause disease.

there aren't any disease -- trick question.

sorry, couldn't resist.

Of course, what is a medically significant variant?

How do you identify and validate those, et cetera?

We have to figure out how we do medical research with large scale medical sequencing

the significance of the data putting clinical with data together et cetera.

As I alluded to earlier, we have to figure how to interact with these patients.

This is a different approach to genetics than we have practiced in the past.

We have to figure out who wants to undergo this medical research, who wants to come to

the NIH clinical center and have to troll through their entire genome and see what we can

find?

An interesting issue.

when I explained this at cocktail parties or gathering I get two die metrically opposed

answers.

One person will say to me how can I sign up for this?

I can't wait.

the other one is I wouldn't let you do that if you paid me a million dollars.

so people have -- feel very strongly about this but we don't understand why that is.

why some people very much want this informationnd others want no part of it?

That's going to have a big impact how we use the tool.

insurance and discrimination is certainly a potential issue.

absolutely.

how --

>> Data release, another good -- our institutes are very keen on large scale data release.

and these data will be released into the public domain in an unassembled manner.

that is the sequence traces will be dumped into JEN bank and will be accessible to all

researchers.

the assembled data and data associated with phenotypes will only be available to

researcher whose have approved studies and gone through a screen.

NCBI set up a database called DB gap.

the database of genotypes and phenotypes which tiered access which is a big hurdle for

NCBI to have come through.

and readers are only able to access these kind of information about they have been

approved and proven legitimate biomedical researchers.

This is the ole fashioned way we have been doing translational research.

start with the phenotype, pick our subjects, dissect them molecularly, then we interpret

and apply those results.

But the question people ask is, okay, you studied this palLISTER hall syndrome or MK

syndrome.

I have a patient exactly like your patient.

but instead of blah blah blah, they have blah blah blah.

could they have a mutation in GLI-3 or whatever?

And I have to always say to them I have no idea.

I didn't study a patient who has what your patient has so I have nothing to offer you.

why is that?

That's because we put patients through this really skinny phenotypic funnel at the outset

and can't answer generalized questions because we have made our study so specific.

so maybe what we need to do is shift to a model of hypothesis generating clinical

research instead of hypothesis answering clinical research.

maybe what we should be doing instead of these narrow phenotypic funnels a collect

patients with broadly defined phenotypes, sort them on their genomic attributes.

Don't sort them on their phenotype it is.

Sort them on their genomics.

Then take subsets of patients who share some genomic attribute and study them in greater

phenotypic detail to -- the phenotypic consequence of a particular genomic attribute?

I thought this is generation.

This is way we normally do it.

we could completely rearrange this paradigm and do this kind of research in very

different ways and it would give us the power to answer questions we just cannot answer

right now.

so here is the question for this audience and for lots of colleagues.

if you had a patient or patients under study and you had complete access to their genomic

sequence, what biological question would you ask?

And I'll bet every one of you could come up with some questions you would love to ask

if you had that kind of a data set sitting on your computer.

here is the other question, what question would your patient ask is that, that's why we're

doing this, to answer the patient's questions about what caused thee things and what they

can do about them.

the future for this kind of research is we are going to generate gigantic amounts of data.

Involved more in bioinformatics than we are in petting and yes going to be totally depen

didn't on transdisciplinary and enteri have a research with colleagues of many different

disciplines to answer these questions that go forward over time.

so I'll stop there.

when -- do you want to do separate questions or joint questions later?

[off mic]



>> Maybe a couple of specific questions.





[off mic]



>> When these peeks are -- you don't know what's causing them so by trolling in and

looking at high penetrant variants you can figure out the gene causing that association.

and vice versa by finding a patient with a rare highly penetrant allele you can generate a

clinical research resource that allows you to test therapeutics and test pathophysiologic

aberrations, it might be hard to test in the low penetrants common alleles.

so it gives you a tool.





[off mic]



>> Do you think it would be possible to identify a totally vanilla person, someone who

has none of these -- no peaks and a normal person?



>> No such thing as normal.

we're all aberrant in our wonderful unique pays.

We knew that from our families.

it's true.

There has been a proposal floated by one of the best genomics people in the country,

Maynard Olesen from university of Washington.

what he say it is phenotype we should be sequencing is longevity.

he says take SENTINARIANS and sequence their genomes.

Those are people who have the attributes -- they have the healthy alleles of lots of genes.

Obviously or else they couldn't have gotten to be 100 years ole.

very provocative approach.



>> My professor of medicine said the value of the stethoscope was that it brought the

patient and the doctor within six inches of one another.





[Laughter]

Speak up there are a lot of people who are listening.



>> You were talking how important it is to get back to your patients with the results.

And I was curious with DB gap now if you start dumping this data and other people are

using it, do you have any hope that you would somehow get reports back if somebody

used that data?

I know this is becoming a big issue.



>> It's a big issue, you're absolutely right.

One of the -- if I remember correctly one qualification for an investigator getting access

to the individual level data and DB gap, they will undertake no effort to recontact the

source of the data.

so they want to not have that circle closed.

you would worry, I would worry and people do the genome wide association studies

would worry they would be buried by well-meaning people reinterpreting their data and

coming up with all kinds of ideas that they would be bombarded with these sorts of things

which none of us could sustain.

so it's an interesting problem, that's how it's been settled.



>> Maybe you want to comment a little bit about translational research other than that's

directed toward inherited disease but more toward inheritance of peoples and migrations

and --

>> Oh.

so not really enough -- I'm not enough of an expert I don't do ethniciography.

I defer to my population genetics experts.



>> We have a group at NIH that actually is involved in that migrations of populations.



>> Not that much.

no.

not that I'm aware of.

I think we should probably -- you had a question.



>> You mention that your only looking at human subjects and that's your focus but seems

like using animal models would be a useful tool after you do identify these things to test

them in animals and do those studies.

I wonderd if you had that component to your project or if you're just reliant on other

people to pick up on those things.



>> Yeah.

so that I haven't figured out how to make the slides three dimensional yet.

but if I could do that my little translational algorithm would be a translation coming out

of the screen of the plane with different species.

You can do that same translational paradigm of malformations and genes with different

species and deduce genes from one species into another.

so for example, we have patients -- we have a very small number of patient whose appear

to have the GCS phenotype.

we can't find -- we have another mouse that looks exactly like the GCPS mouse but

doesn't have a mutation in GLI-3.

So that mouse is a model for the humans that don't have that mutation.

so we're positionnally cloning the gene in the mouse because the humans are too rare to

do it.

Find the gene in the mouse and test the humans so we'll close the circle in an interspecies

way doing exactly that.

absolutely.



>> So Les, where do most of your patients come from, in the sense is it a physician who

calls up, is it the family who calls up?

Are the patients more ahead of the doctors in terms of accessibility to, I don't know,

Internet, to even awareness of --

>> Two of the phenotypes we taked about today the frequency of some of these --



[off mic]



>> Oh, sorry.

he's waving at me.

everywhere.

so I'll stop there and let Julie go.



>> And they're all brought here at public expense and housed and everything?



>> We saw mixture of a little (inaudible).



>> Okay.

Julie.



>> Great.

can everybody hear me?

I'm going to wait to see if anybody comes out et and tell me the mic isn't working.

so I'll spend time talking about a concept at the intersection of these ideas of scientific

exploration and where are the interfaces of that where w the family.

I'll talk about genetic diagnosis today.

and as he mentioned I'm a genetic counselor and since many of you are maybe not

familiar with this particular profession, I'm going to go into a little more detail about

what exactly genetic counseling is.

sometimes I think when you're reading a paper a manuscript you'll come across a

sentence that reads something like the patient received genetic counseling about the

disorder and about their reproductive risk and they demonstrated good understanding.

I think that's a simplistic view of genetic counseling and doesn't encompass all of it.

I think one of the official definitions in genetic counseling sit's the process of helping

people understand and adapt to the medical psychological and familial implications of

genetic contributions to disease.

When I think about what looks like in the clinic I think about communication, it looks

like a communication process.

so genetic counselors communicate with their patients about things like genetics and

biology, inherry tense of genetic disorders, what kind of testing options are available and

how to manage these conditions in the family.

sometimes when they're called upon to facilitate decision making whether or not to

undergo particular genetic test or what to do with that information is certainly something

within the purview of genetic counseling.

I think most of the time it's very important to assess the different psychological factors

and issues that can come up for families experiencing these -- undergoing genetic

diagnosis.

and so it's usually a short term interaction driven by the agenda and the concerns of the

patient or client, what their questions are and what their goals and expectations are of the

session.

and genetic counseling although it exists as a discrete profession, many people engage in

this.

so people varying from clinicians and geneticists to nurses and social workers like I said,

it does exist as a discrete profession but there are lots of folk whose do this on day-to-day

basis and is an integral part of their practice.

uh-huh.





[off mic]



>> Uh-huh.





[off mic]



>> So the question was are -- can genetic counselors become board certified, who is the

organization that certifies them?

And the organization in the United States is the American board of genetic counseling.

it's a discrete organization from the American board of medical genetics though they used

to -- there was a period of time where genetic counselors were certify bid the medical

genetics organization.

in Canada they have their own certification board.

so genetic counseling usually starts with diagnosis of genetic disorder or the

consideration of a genetic disorder.

so today what I hope to do is give you an overview of the concept of genetic disorder

from this perspective from the perspective of genetic counseling.

I hope that will tell you more about what genetic counseling is than any definition could.

so I think that definitions are a good place to start.

and when I looked up the word diagnosis in Miriam Webster it's defined as the art or act

of identifying a disease from its signs and symptoms.

And so even though I'm going to mention some of the techniques that are employed in

the laboratory to make these types of diagnoses, I'm going to focus less on the act of

obtaining molecular diagnosis and think more about what I consider to be the art of

genetic diagnosis.

And when I say the art of molecular and genetic diagnosis I mean thinking about this as

more of a complex and iterative process.

rather than a single discrete identification of a sequence variant in the laboratory or by the

recognition of a malformation syndrome by an astute clinician, thinking about it more as

a process that involves over -- evolves over time rather than a single discrete thing that

happens just in one setting.

so my goal today is to use a few clinical vignettes to illustrate how diagnose affects

families with genetic condition.

so a genetic diagnosis has a lot of things in common with diagnosis of other things like

communicable disease or things like that.

It provides an explanation for an underlying medical condition or concern.

it places some boundaries object condition and gives people an expectation of what to

expect for the future.

prognosis.

and it reduces uncertainty, a lot of times family whose are seeking a diagnosis whether

genetic or otherwise are wondering what's going on?

It puts a label on the condition.

of course this label is not usually the first time the family has an idea they there's

something going on.

this is the first time that it's labeled for them.

however, I do think it's fair to say diagnosis of a genetic disorder has a number of

characteristics that set it apart from a diagnosis of any other kind of condition.

it's these differences that I think are worth keeping in mind as I present cases.

so genetic conditions are generally rare.

they don't affect a lot of peep in the population.

These are mendelian disorders and I'm going to spend most of my time talking about.

and I think a lot of times people wonder what is the utility of having a diagnosis of this

kind because it's permanent.

they aren't curable in the sense that many other conditions are.

Of course when we think about genetics we always have to think about the fact that these

things are heritable.

they're considered to be intrinsic to a person and often times on a molecular level these

changes are unique.

my goal here is to kind of talk about some of these difference, highlight some of these

differences and how we think about them in the clinic.

so I'll move right into a case.

This young man has a diagnosis of lens microthat willMIA syndrome.

the main feature is evident in these pictures.

You can't hear me when I move that way.

He has he was born with very small or absent eyes.

has a number of other physical differences that are not evident in these pictures.

and I think that just to give a quick clinical synopsis, it's characterized by a number of

different eye mall formations ranging from microPHTHALMIA, a number of limb or

skeletal anomalies so patients can have a number of anomalies or differences in digits.

Usually they're small for their age.

and there can be a number of urogenital anomalies that happen in these boys.

they're almost always cognitively impaired mentally retarded or developmentally

delayed.

and this is a disorder that is characterized by X link inheritance. There's two loci on the

X chromosome associated with this particular condition.

so in clinical genetics we usually draw a diagram or graphic representation of how a

disorder is segregatening a family called the pedigree. For those not familiar with this,

the Ohio familiar the squares are men and the colors are women.

we draw an air reto the first person in the family affected with condition.

so this is a pedigree of a typical family with lens syndrome.

When we take a investigation of trying to determine the mutation in the family we find

this, that the child has a copy of -- has a mutation in a gene in one of these genes that are

associated with lens syndrome.

and that his mother is a carrier for that gene.

of course the mutation status of the two of his two sisters is unknown but since the mom

is a carrier for this condition, it's possible for her -- each of her unaffected daughters to

have inherited the gene.

the boy himself is affected the mother is not affected because she has two copies of that

particular gene though one has a mutation.

The other -- the gene on the other X chromosome is acting like a back up copy so she

doesn't show signs of the condition nor do the daughter bus her daughters are at risk of

having infected sons.

Does that make sense to everybody?

Okay.

so when we look at the family, a number of different things -- a diagnosis can have a

number of effects on the family.

so people might wonder -- mom might wonder about her sisters if they're carriers of gene

and whether or not their children are at risk to have the same condition as her son.

I'll tell you a little bit of a story this family.

the affected boy was conceived using in vitro fertilization.

His mom was convinced the reason why he had lens syndrome was because he was

conceived using assisted doctor technologies, she thought his birth defects were caused

by the manipulation of egg and sperm outside the body.

and she also thought that since she and her husband had together decided to use in vitro

fertilization as their technique to get pregnant that the blame was evenly split between the

two of them.

when they received this diagnosis and received the information that there was actually a

mutation in the gene that was responsible for causing the phenotype this was -- caused

her to have a number of different feelings.

and she was actually really upset by this.

it made her take on a lot of guilt about this condition.

she really felt like this was absolutely something that was her fault and she also felt like

her husband would blame her now and that more of the blame laid on her shoulders.

and immediately she recognized the significance for her other children.

so she didn't think her daughters would be at risk because she hoped they would be fertile

and not have to use IVF to get pregnant.

now she knew there was a chance her daughters might have also had this mutation.

So in the counseling session with the family we were able to identify all of these concerns

and we spent a lot of time talking about them.

so I watched this mom have -- go from a person who had really thought that the reason

why her child was affected was this condition was something out of her control to

something that was completely within her control.

And that she felt very guilty for this.

she had a heightened awarness for family history and genetics and became very interested

in it and we wound up talking about it for a really long time.

so this family I think is really a good example of what I think of as a family with an X

link disorder and how the thoughts of guilt and blame play out in these families.

Parental guilt and blame are absolutely factors to consider in all -- whenever you're

talking with the family who has a genetic condition, but I think for this family, they

represented a really good example of the idea of causal attribution, basically how a

family answers the question, why did this happen or how did this happen?

And I think when we think about molecular diagnostics, we can answer the questions of

how it happened.

This happened because the mutated gene and caused a different set of instructions to get -

- set of instructions to get sent to the body but it doesn't help for people to answer the

question why did this happen and why to me and why was it my family that this had to

happen to.

answering that question and coming to terms with the which your family answers that

question is really one of the first steps on the process of adapting to having a genetic

condition in the family.

something we spent a long time talking about with this family.

so we've recently -- a lot of people who work in clinical genetics have had this sense that

guilt in X link disorders or guilt in moms with children with X link disorders was more

apparent than in conditions that are inherited in other ways.

Some recent work has come out to support that.

so moms who have boys with a X link disorders do have an increased awareness of

genetics and how these alleles segregate through families.

They worry a lot about their female relatives.

they perceive more blame from their husband ant partner than perhaps in an autosomal

position where both parents contribute an altered gene to produce the child's phenotype

and we know unresolved guilt leads to poor psychological outcomes so it's important to

address these issues and we work -- when we work with these families.

so this I think is a great case to illustrate some of the psychological sequella that can

come from genetic diagnosis.

We talked about guilt and responsibility.

this is something that lots of families cope with.

the idea of blame, who exactly is responsible for this, is absolutely something that is an

essential thing for families to talk about and to consider.

can cause a shift in family dynamic, patients who are considering a genetic diagnosis, I'm

going the talk -- actually present a case to illustrate this point a little bit more later on.

can describe how things were different in their family after this information came to light.

these sound like bad thing and a lot of times receiving this information is not good news.

on the other hand sometimes getting a genetic diagnosis can instill a sense of relief

because this process of trying to find out what is wrong with the child is finally over.

they have an answer.

and of course that helps people to get more control over the situation.

and feel like they're more in control of the situation.

so the second case that I'm going to present many of you may have an idea what the

diagnosis is already by some of the pictures that Dr. Biesecker show earlier.

this little bit is about two years old and he has facial differences and some physical

differences that were noticeable shortly after birth.

His head is larger than you would expect for a child his age.

his eyes are widely spaced.

He has developmental delay, isn't accomplishing the same milestones that you expect a

toddler his age to accomplish.

He was born with two -- with extra big toes on both feet and this picture is after he's had

surgery to remove those.

And he can also see the webbing between the second and third toe on each foot.

so he's the only child in his family with any features that are even remotely close to this.

there's no history of polydakCTILY developmental delay.

so the two points I want the make for this family, this child before he was seen in our

clinic went through at least three previous diagnoses prior to being diagnosed with this

entity that Dr. Biesecker talked about, the the GCPS CGS with a technology caused array

CGH.

let's tackle the first point first.

this child had received at least three previous diagnoses, we spent a long time talking

with the family about it.

here he is two years ole and they're getting the answer to what's going on with him.

they described a process of being on what we call the diagnostic odyssey.

I think odyssey is great word to use here.

Families describe this process of coming to a diagnosis as being extremely exhausting

and it really does seem like they're on a journey that seems like it will never end.

so it's defined as the search for a diagnosis or explanation for their child's condition.

and it's very common in instances of rare conditions.

simply because they're rare there aren't a lot of clinics familiar with what's going on and

sometimes if you don't know the test to order or if it's unavailable it can go on a long

time.

it's emotionally and physically exhausting.

Takes a real toll on the family to continue to poor resources emotional, physical financial

resources and trying to get an opinion on what's going on with their child just take as lot

out of them.

and I think that one thing a lot of families describe to us are feelings of responsibility.

When a family talks with their neighbors, family members, what's going on with your

child, did they find a diagnosis for your little boy yet?

A lot of times parents perceive lot of blame when they aren't able to provide an answer

for that. It implies they aren't a good parent, that they haven't accessed the right medical

resource, they haven't made enough effort into trying to investigate what is the cause of

their child's condition, can be tremendously stigmatizing for family,s them to be anxious

and it really ups the uncertainty in the situation.

because each diagnosis when it gets put into place and gets take away seems to be worse

than the last and they don't have any frame of reference or boundaries on their child's

condition.

they don't have a good idea what to expect, what the prognosis might be so it leaves them

in a situation of tremendous uncertainty, not a good place to be.

and finally a lot of families describe that being in the undiagnosed state makes them feel

like they don't have a community.

their friends, neighbors who have children with Downs syndrome or autism, they have a

place to go for their support group.

they have a place to get resources and to connect with other families and with other

parents.

but parents of children who don't have a diagnosis feel like they just don't have this

community and oftentimes feel like this leads to a lack of resources so they can't get the

physical or speech therapy because no one has a diagnosis for the child approximate it's

harder to access things in that ininstance.

the second point to make from this case is how this child was diagnosed with this

contiguous gene syndrome with this technology called array GCH.

let me review it quickly, this is a condition characterized by microreceively, a bulging

Ford head, widely spaced eyes and autosomal dominant inheritance which means if a

person is infected each child has 50% of inheriting the GLI-3 gene and developing the

condition.

this sin chrome also presents with things like developmental delay and seizures.

array CGH to quickly review is a technology where we basically mix patient DNA and

control DNA that has been labeled together and allow them to hybridize to an array.

at the end of the day this allows us to see a gain or loss of genetic material.

to basically detect copy number variation.

so this child has a large deletion.

on chromosome 7.

we can see the GLI-3 gene is there.

but the question that we have and the question that the parents have are what about all

these other genes?

How do they contribute to the Phenotype?

This is a very interesting scientific question and also an interesting question for this

family.

They would like to know more about that.

and my point here is that a diagnosis can evolve over time.

rare disorders are infrequently completely delineated an understood so as we learn more

about the different contributions in this example, that the difference genes in the deleted

region make and contribute to the Phenotype will have more to tell the family technology

has become available, we've learned exactly what genes they're deleted for, we know the

consequences of having being delete Ford those genes and we're calling them back and

telling them the child's disorder is different, it's a broader phenotype than what they had

originally thought.

that's an interesting process for the family to undergo.

that means there's conning education essential both on the part of the clinician and family

feels like they need to become experts in this so they can learn how their child is -- child's

diagnosis may change and how they get used to that over time.

how do they get used to those changes over time.

so the last point I want to make is basically a complicated point about family systems.

just like a genetic diagnosis evolves over time the percentage of genetic diagnosis can

change over time with a family.

this family I'll property right now, there's their pedigree, they present -- pedigree.

they presented all the people colored presented with two features.

They had a epiglottis and history of post axil polydakTILY.

This was identified as a genetic syndrome when the man in the second generation went to

see the ire nose and throat doctor and the physician detected it and he was intrigued by

this and asked the man were you born with extra digits?

And he said I was born with extra pinky fingers on both hands and so was my mom and

brother and so this pedigree was emerged and he was referred to genetics to confirm the

diagnose know so I was palLISTER hall syndrome.

When we discussed it with them we were hesitant on whether to give them the diagnosis

of it because we performed MRIs and none of them had a hyperTORAMIC hemotoe ma.

-- none of these individuals had a high toe ma so it was interesting to talk to them about

whether or not when we undertook this examination of the GLI-3 gene to see if weld

uncover genes in that gene and indeed we did.

now we are faced with a situation where the family molecularly could be given a

diagnosis of palLISTER hall syndrome because they have a GLI-3 alteration but they're

mild on that end of the spectrum.

what's also interesting we performed routine caryo types offer everyone in the family. It's

a test that looks at a person's chromosome.

and what we found actually was that there was a familial chromosome translocation also

segregating at the family, has nothing to do with a mutation so this family essentially has

two things.

They have a mutation and a gene and they also have a chromosome translocation.

this was another very interesting discussion to have with the family.

as one of the family members put it, it seems like our family has something.

and if you look at their family it's true.

Everyone in the family has either has the translocation or has this polydakTILY

phenotype or -- yeah, or both.

and several people do have both.

now one of the consequences of having a chromosome translocation is it can increase the

person's chance of having a miscarriage.

and so this was another piece of information that I think the family wasn't really ever

intending to learn when they started talking with us.

as our conversations with the family evolved several people in the third generation

reacted differently to this news.

A lot of these children were enrolled in our study when they were quite young and by the

time we were able to accumulate all this data they were older.

and so they had originally understood that we were embarking on this investigation to

identify the cause of their family trait.

now they have a cause for their family trait.

it may have the word syndrome attached to the end of it.

and then they also may have this chromosome translocation which could increase their

chances for miscarriage.

several members of the family had a really hard time communicating about this and some

of the spouse whose had married into this family went on the Internet looked up the

syndrome an had the impression that all the children were going to be born with mental

retardation and just tearial birth defects.

it was a big -- it was kind of a mess actually.

and we had to do a lot of talking with each of the different people in the family.

and I think this is a really good illustration of why it's important to take into account the

family system when we're giving people these diagnoses.

genetic conditions affect families as well as individuals.

and the family dynamic can be seriously affected by the diagnosis of a genetic disorder or

a genetic pre-disposition.

sometime there's a lot of secrecy surrounding things like non-paternity, misattributed

paternity in a family.

these are absolutely things that can come out when you're investigating the cause of

genetic condition.

it also brings to light inherent variations in the way people cope with this information and

how they communicate about this information.

this is especially true in some families that we see in clinic, not me personally but a lot of

genetic counselors work with families who have inherited cancer predisposition

syndromes.

Various members of the family some family members may not be interested at all

understood undergoing genetic testing.

others are extremely interested.

Some families right in front of you, these kinds of philosophies can colied and it'

important to foe gaucheiate these things and help patients go into discussions with their

eyes open, that maybe people that don't feel like you do undertaking this kind of

investigation and it's something you might need to negotiate later on.

I'll conclude by mentioning something that is something that we have to discuss with our

patients almost all the time.

This is idea of the number of misconceptions when we talk about genetic diagnoses.

I don't know if anyone has seen the movie the incredibles the Disney movie.

the nemesis is actually named syndrome.

it wasn't really -- I wasn't that happy to see that.

it has an incredibly negative connotation in our society.

most people equate the word syndrome with mental retardation.

we just presented here a number of families can have some of these syndromes which can

be as mild as inherited family traits completely cognitively intact so when people

sometimes hear that word they get very upset.

there's a lot of stigma attached to the idea of having a genetic diagnosis.

people really worry about genetic discrimination and how this information can be

misused by employers, by insurance companies, and then there's this whole idea of

genetic determinants that we're constantly fighting against that we cannot -- that our

future is entirely determined by our DNA.

so I think that in closing isle just mention that the ethical, legal and social issues are

absolutely essential to investigate when we think about this concept of genetic diagnosis.

so to summarize, genetic diagnosis is complex and the importance of molecular diagnosis

varies from family to family.

I think the reaction Dr. Biesecker described going to a cocktail party and telling people

about undergoing these kinds of investigations is absolutely the same thing with different

family members.

For some a families understanding the molecular cause of their child's disorder is

incredibly important.

and they really appreciate every ounce of time that goes into explaining how those things

happen and understanding these types of results and what happens in the lab and how we

arrive at this information.

And other people, honestly it doesn't matter to them.

they would really just much rather focus on what the prognosis is for their child and

coping with the cards they have been dealt essentially.

so the careful consideration of the psychological and psychosocial impact is absolutely

essential when we consider genetic diagnosis.

That's I think the main theme of my talk.

all of this work is absolutely impossible without a number of different people.

even when a family is of the opinion that understanding the molecular basis of their

condition is not important we think it's very important and we really appreciate all the

work that different members of our lab do to allow us to make that possible.

and of course, without the patients and their families, none of the work that we do would

be possible either.

so we both would like to thank them also.





[Applause]



>> Thank you very much.

that was very exciting.



>> Question.





[off mic]



>> Secrecy, non-paternity.

I didn't catch the connection between secrecy and non-paternity.



>> So oftentimes families keep things like non-paternity a big secret.

and occasionally when we invest --



[off mic]



>> So when the social father is not the biological father.

and so this can absolutely come out in the course of investigating the cause of a genetic

condition.

does that make sense in you seem like you were about to ask a second question.



>> The father in the family is not the biological father.



>> Exactly.



>> And that might be a secret that the parents keeping or that the mother is keeping.



>> A good example -- exactly.

so a good example might be an autosomal dominant disorder where a child and uncle,

they are affected but the social father of that child is not affected.

and we identify the mutation in the child, we find the same mutation is in the uncle,

maybe the uncle is the father of the child and that hadn't been known beforehand, the

father doesn't have the mutation.

does ma make sense?

So it's a big issue.



>> So in terms of sort of the -- your approach here at NIH, you gave us an example of

hypercholesterolEMIA, a common disorder and how you would go about it.

but do you establish certain entry points into a study or do people call more like the first

patient who you describe where there's some abnormal phenotype and other people in the

family have it?

How do you decide who to accept and what topics because it's so vast.



>> It's the scientific equivalent of attention deficit disorder.

we have that.

we spend every Tuesday Julie and I meet to consider all the inquiries that we've received

and decide who is eligible for the studies and what priority level do those people have.

we have a limited amount of resources and clinic time and our time, et cetera.

we have to prioritize.

research opportunities, we do try to be somewhat respectful of some families can be stuck

in problematic circumstances that we might be one of the few or only people that can

untangle for them.

sometimes we'll make a decision to bring a family in more for compassion.

it's a really hard thing to do and there's much more need than there is resource to address

it.

That's for sure.





[off mic]

In

>> It actually stands for GLIOMA associated Congress gene, it was originally identified

as being amplified somatically in glioma tumors which then turned out to be false

because in those gliomas it was a neighboring gene, MDM-2, that was in fact amplified

in the tumor, not GLI-3 as a causative aberration in those tumors but turns out GLI-3 is a

member downstream of the same pathway that MDM-2 is.

in fact some patients with these phenotypes are susceptible to tumors.

Many wrong steps along the way but turns out to be related.





[off mic]



>> Julie.



>> What is the percentage of miscarriages in woman?

I was I was told it was about 50%.

it takes care of all the issues and don't want to spend the time (indiscernible) so with the

problem of you say --

>> For the translocation.

so it varies depending on the translocation and the amount of material that's actually in

the wrong place.

so for some chromosome translocations it can be extremely difficult to maintain -- the

probably is quite high.

for other translocations the probably is quite low.

depends.





[off mic]



>> I'm sorry --

>> The question was with a particular familial chromosome translocation what are the

chance force a miscarriage?

It varies depending on which chromosomes are involved and which chromosomes -- the

extent of the translocation.

Another risk that people consider when considering risks of carrying a familiar

chromosome translocation are risk to have abnormal live born offspring so live birth with

significant abnormalities because of miss organize extra genetic material.



>> I was interested in your difficult decision in the family that had the mile case of the

palLISTER hall syndrome and why you decide in the end to give him that name rather

than say you have a mutation in such a such protein.

>> It really -- the way it actually played out in our discussion was that we discussed this

issue extremely openly with them.

we talked to them about how when we think of a clinical diagnosis of palLISTER hall

syndrome we think of a triad of anomalies.

The truth was clinically they didn't seem to meet that description.

but molecularly, they absolutely had this mutation that was consistent with a clinical

diagnosis of palLISTER hall syndrome.

so in our discussions we explained this to them.

some chose to think of this as the syndrome and think of this as being a mild case ofP

palLISTER hall syndrome.

they had a molecular explanation for family trait so it really just came down to an

individual preference for the family, who was most interested in how to refer to what had

started off as a family trait that many people didn't even remember they had.

All the children in the third generation who had their extra digits removed before they

were able to form memories so they weren't aware that they had ever had this trait.

so it turns into a discussion of how should we label this.

for some people it was easier to think of it has the milest case ever palLISTER hall

syndrome and we probably couldn't even call you a patient with palLISTER hall

syndrome and others chose to think about it in molecular terms, we have a molecular

explanation for a family trait.



>> you mentioned you talked to the entire family and people that married into the family.



>> Uh-huh.



>> Did you you can to them as individuals or as couples or ooze entire family and how

do you decide how to do that?



>> All of the above.



>> So how do you decide how to talk to them?

Or do you just do every combination and be done with it?



>> We usually approach people -- before -- this family is nowhere near from here.

so they all traveled here to meet was so that gave us an opportunity to touch base with

them over the phone before coming here.

so I usually touch base with people on an individual basis and ask them would you like

for your parents or your sister or brother to be involved for the discussion, how would

y'all like for us to approach this?

And most families would like to hear one explanation of genetics.

so you don't repeat yourself so that everybody hear the same thing.

and they would like some victimized attention.

there's the opportunity to be recontacted by phone.

so -- right.

so much of this counseling took place over the phone.

another issue -- hurdle that we have to overcome is privacy.

we need explicit permission from each family member to discuss another family

member's situation or results, we can't even acknowledge they're our patient unless they

have given permission.

Sometimes it's easier, sometimes it's hard tore have a conversation with someone talking

about hypothetical diagnosis.



>> Here is a thought that maybe off the wall.

maybe not appropriate.

most of the phenotypes you discussed are physical attributes.

But nothing about behavior or mental attributes.

Is this an area where you want to avoid -- say for example criminal tendency or

something like that.

that's linked to the other phenotypecal thing?

Just a thought.



>> Psychiatric genetics is a huge area of current research that is underway in the NIH

primarily in the national institute of mental health appropriately enough and lots of

people are working very hard to identify susceptibility genes for psychiatric diseases.





[off mic]



>> Oh, yeah.



>> Degeneration did you say?



>> Related issues.

it's been a tough, tough problem to crack because the psychiatrists aren't -- don't have

good luck with Phenotyping.

I can count fingers and I can measure these things in people's brains and those are

unambiguous objective phenotypic data and distinguishing between severe bipolar illness

and mild schizophrenia is a very hard thing to do and most of the pedigrees multiplex

families with psychiatric disease are let owe genius in -- heterogenous in phenotype and it

is a tough problem to solve.

so they have been working on it as long -- they have been working on the psychiatric

phenotypes as long as we have been working on the physical phenotype, we have made

more progress not because we're better geneticists but because our phenotypes are more

amenable to this process.

so it's just going to take a long time to get the psychiatric phenotypes through.





[off mic]



>> These psychiatric studies, they have enrolled thousands of people.

There's plenty of motivation on the family's part to untangle this, it's just a tough

biological clinical phenotypic problem.

>> I think the jury is still out as to whether or not having a genetic or biological

explanation for a particular behavior will actually go a long way towards reducing stigma

or whether or not the stigma of the particular disorders will remain the same.

that's another concern the families have but hopefully coming up with a biological

explanation will reduce the stigma associated with some of those disorders.



>> I'm curious, do you ever record your sessions with the patients, with the families, and

then give them a CD?

The reason being that it's well known in medicine that when parents are under stress and

fearful of what the disease is, you know, they hear the loaded words and then two weeks

later or maybe two days later when they're explaining to somebody else in the family they

make another soup about it and everybody gets all riled up.

so that's always been a very inefficient way I thought of communicating information.

Giving it to the doctors with all due respect often isn't any different, particularly in this

new world you're talking about which raises another question, how are you going to get

physicians of the future to be capable of not only approaching medicine in this light but

taking advantage of who is going to translate the information to operational, maybe you'll

tell us that one too.



>> I'll answer the first part.

so to address the first question there's some genetic counseling students in the audience

and they kind of laugh because in training it's quite common for genetic counseling

students to audio record their sessions.

do we ever transcribe those sessions in a formal --



[off mic]



>> Well it's to help in their training.

They record the sessions to help them in their training so they can review -- exactly.

right.

Right.

so your question --



[off mic]



>> Exactly.





[off mic]



>> So we don't -- we make those recordings for the students benefits not for the patients

benefits.

I think that's something that in our group we spend a lot of time talking about is what --

and we're not the only people talking about this but what is the most effective way the

communicate these complex topics genetics, and of course all the psychosocial sequella

of having this information to patients.

We currently write our patients letters that we hope communicate adequately the content

of our discussion and the Tenner of our discussion but there are many people working on

that exact question.



>> It would be interesting.



>> It would be an interesting idea.



>> Studies in medicine for this exact question was examined and always turned out it was

better to give the patient an audiotape where you hear the voices and they remember the

whole business.

but it was a matter of expense.

nobody ever did it.



>> These letters, worth emphasizing the typical patient receives within two to four weeks

of being here usually a three to five-page single spaced letter from Julie that delineates

everything that all evaluations that we did while they were here, the testing that they

underwent, a discussion of the inheritance pattern of the disorder they have, the range of

severity of disorder, the interventions that are appropriate and the recommendations for

future.

it's a gigantic amount of work which is why they argue about how much -- how you

spelled it but one thing interesting that we've heard that the DOCS, the referring doctors

like her letters bear than my dictation.

because her letters are of course written in English.

mine are written in medicalEASE so hers make more sense.

so DOCS like those better.

interesting.

so your question was on diffusion knowledge.

medicine is an inherent literbly conservative profession.

things change very slowly which is in some ways is good because it prevents patients

from being subjected to the whim of the moment.

that's buffered patients from a lot of disaster over the years, a lot of dumb things people

wanted to do.

but one thing it is not good at as rapidly responding when the science and technology

changes, it take as long time for these things to diffuse out into the practice community.

however, that being said, the criticism of us by the medical community is that we don't

produce enough stuff that is directly useful to them in their every day practice.

that's their criticism.

and they say we'll be happy to adopt your gee whiz technology as soon as you can show

me that it will change.

What I'm doing to my patients in my clinic today.

that's really the challenge to us is to try and again turn some of these discoveries into

things that actually work in the trenches on the ground today that the doctors need to take

care of their real world patients.

>> How much does it cost to have your genome sequenced as a private individual?



>> Well there's two companies now that are offering high through put genotyping, they'll

do hundred thousand genotypes or half a million genotypes the utility of that is quite a bit

under debate.

put it politely that way.

there is a couple of companies that are about to launch sequencing private for hire

sequencing.

the prices I'm not aware have actually been set yet.

I expect they'll be in the range of order of magnitude range of about a million dollars for

the first people who uptake this.





[off mic]



>> For the companies.

But again as soon as these new machines become effective it will probably start

dropping.





[off mic]



>> The question is do the sequencing pick up some of these repetitive elements.

so depending on the technology that you use it can.

some will, some won't.

sequencing also does not -- is not terribly good today at picking up copper -- copy

number variations, duplication deletion polymorphisms that are large in size which we

now know there are thousands of them in the genome.

it doesn't pick up epigenetic changes such as imprinting changes so it's not as though the

sequence is all of biology, none of us think that.

but it's a lot of biology.

if we could get the sequence we could do a lot of things, doesn't mean it's everything.



>> Okay.

I think we should let our guests go.

thank you very much.





[Applause]