GENOME-WIDE LINKAGE SCAN IN A LARGE FAMILY WITH SUSPECTED GENETIC
Document Sample


GENOME-WIDE LINKAGE SCAN IN A LARGE
FAMILY WITH SUSPECTED GENETIC
PREDISPOSITION TO CANCER
Marroni F.1, Lipkin S.2, Benatti P.3, Pedroni M.3,
Bailey-Wilson J.E.4, Collins F.4, Ponz de Leon
M.3 and Presciuttini S.1,4
1 Università di Pisa, 2 University of California, 3 Università di Modena e Reggio
Emilia, 4 National Human Genome Research Institute, Baltimore
Genome wide linkage analysis
• 392 markers ( 10cM average distance)
• 104 subjects, of which:
– 31 completely typed
– 9 partially typed
– 54 untyped
• Three point linkage analysis using GAS and FASTLINK
• Parametric and non-parametric multipoint linkage analysis
using GENEHUNTER
Genome scan results
• A peak of lodscore 2.0 was obtained for
two adjacent markers on chromosome 17
Genome scan results
3
2
Lod Score
1
0
-1
-2
0 20 40 60 80 100 120
Map position [cM]
Further studies on chromosome 17
• New sampling campaign: we obtained DNA
samples from 10 new subjects
• Of them, one gave rise to mendelian
inconsistencies and was therefore discarded
• For chromosome 17, typed subjects increased to
46 (since three of the nine partially typed were not
typed for Chrom17)
• In addition, three new markers were typed
near the peak region
Further studies on chromosome 17
3
2
Lod Score
1
0
-1
-2
0 20 40 60 80 100 120
Map position [cM]
Genome scan results
• 2 more interesting regions, with peaks of
lodscore of 1.0
1.5
1
Lod Score
0.5
0
-0.5
-1
-1.5
0 50 100 150
Map position (cM)
Genome scan results
• 2 more interesting regions, with peaks of
lodscore of 1.0
1.5
1
Lod Score
0.5
0
-0.5
-1
-1.5
0 50 100 150
Map position (cM)
Power analysis for further studies
• We chose a real marker from one of the
interesting regions
• 9 alleles: H=0.74
• In order to be able to run Fast-S-Link we
had to renumber them to 6 (H=0.736)
Simulation results
1.0
0.8
0.6
Mean Lod Score
0.843
0.4
0.462 Linkage
0.2 No linkage
0.0
-0.359
-0.2 -0.587
-0.4
-0.6
Original Extended
Simulation results
2.5
2
Mean Lod Score
1.5
1
0.5
0
Real Real+NewID Real+All the living
Conclusions
• We may obtain a substantial power increase
for linkage analysis by adding to our study
the newly sampled subjects
• This suggests that a possible option is to
extend the typing of the entire marker set to
the new subjects
Get documents about "