Comparison

Document Sample
Comparison Powered By Docstoc
					Analysis of UV sensitivity

   To compare results obtained with our method to those obtained using a competitive
growth/ barcode Affymetrix hybridization method (Winzeler, E.A. et al. (1999) Science
285, 901-6), we examined the results of a published study of UV sensitivity (Birrell,
G.W., et al. (2001) Proc Natl Acad Sci U S A 98, 12608-13). This study was chosen
because utilized the same collection of deletion mutants, the homozygous deletion
strains, and applied similar levels of UV irradiation. Of 40 non-essential genes with UV
sensitive phenotypes previously described in the literature, the Birrell study identified 24
(Supplementary Table 1). Using our conservative criterion, mutants displaying growth
defects in both experimental replicates, we identified 13 of these mutants. By the more
relaxed criterion of mutants displaying growth defects in only one experimental replicate,
an additional 6 of these mutants would have been identified (Supplementary Table 1).


            Gene           ORF           Reported        Conservative       Relaxed
         RAD2           YGR258C            High             strong           strong
         RAD10          YML095C            High             strong           strong
         RAD14          YMR201C            High             strong           strong
         RAD5           YLR032W            High             strong           strong
         RAD18          YCR066W            High             strong           strong
         RAD7           YJR052W          Moderate           strong           strong
         RAD9           YDR217C            High           moderate         moderate
         HPR5           YJL092W            High           moderate         moderate
         DUN1           YDL101C            Mild           moderate         moderate
         RAD23          YEL037C          Moderate         moderate           strong
         MEC3           YLR288C          Moderate         moderate           strong
         RAD24          YER173W          Moderate         moderate           strong
         RAD17          YOR368W          Moderate         moderate         moderate
         RAD1           YPL022W            High          No phenotype        strong
         MMS2           YGL087C          Moderate        No phenotype        strong
         RAD4           YER162C            High          No phenotype      moderate
         RAD57          YDR004W            Mild          No phenotype      moderate
         RAD55          YDR076W            Mild          No phenotype      moderate
         RAD51          YER095W            Mild          No phenotype      moderate
         RAD59          YDL059C            Mild          No phenotype     No phenotype
         SAE2           YGL175C            Mild          No phenotype     No phenotype
         RAD54          YGL163C            Mild          No phenotype     No phenotype
         MUS81          YDR386W          Moderate        No phenotype     No phenotype
         REV7           YIL139C          Moderate        No phenotype     No phenotype

         Supplemental Table 1. UV sensitive mutants described in the literature identified
         by Birrell, et al.
         Gene           ORF         Birrell            Conserv.            Relaxed
         RAD6           YGL058W     Low Hyb            strong              strong
         UBC13          YDR092W     Low Hyb            moderate            strong
         XRS2           YDR369C     Low Hyb            moderate            moderate
         DDC1           YPL194W     Low Hyb              No phenotype      moderate
         MRE11          YMR224C     Low Hyb              No phenotype       No phenotype
         RAD50          YNL250W     Low Hyb              No phenotype       No phenotype
         RAD52          YML032C     Low Hyb              No phenotype       No phenotype
         MET18          YIL128W     Low Hyb              No phenotype       No phenotype
         REV1           YOR346W       No phenotype       No phenotype       No phenotype
         REV3           YPL167C       No phenotype       No phenotype       No phenotype
         RAD16          YBR114W       No phenotype       No phenotype       No phenotype
         RAD30          YDR419W       No phenotype       No phenotype       No phenotype
         CAC2           YML102W       No phenotype       No phenotype       No phenotype
         RLF2           YPR018W       No phenotype       No phenotype       No phenotype
         UMP1           YBR173C       No phenotype       No phenotype       No phenotype
         MSI1           YBR195C       No phenotype       No phenotype       No phenotype

         Supplemental Table 2. UV sensitive mutants from the literature that were not
         identified in the Birrell study.



   Of the remaining 5 mutants identified by both the literature and Birrell, et al., 3 were
reported to possess only mild UV sensitivity, possibly highlighting the greater sensitivity
of the competitive growth method. In contrast, our conservative criterion detected 3
mutants with phenotypes reported in the literature that were missed in the Birrell study
due to poor barcode hybridization (Supplementary Table 2), highlighting an advantage of
our plate-based method. The fact that both the Birrell study and our study failed to detect
phenotypes for 12 mutants described as UV sensitive in the literature (Supplementary
Table 2) suggests the existence of strain background differences or errors in the yeast
deletion set. Finally, our results confirmed the UV sensitivity for 2 of the 6 novel UV
sensitive mutants identified in the Birrell study (Supplementary Table 3).


                  Gene           ORF         Conservative          Relaxed
               MMS4           YBR098W        No phenotype        No phenotype
               YBR099C        YBR099C        No phenotype        No phenotype
               THR1           YHR025W       Not in collection   Not in collection
               LSM1           YJL124C        No phenotype        No phenotype
               SNF2           YOR290C            strong              strong
               YAF9           YNL107W          moderate           moderate
               YHR134W        YHR134W        No phenotype        No phenotype

               Supplemental Table 3. New UV sensitive mutants identified
               in this study.
    Taken together, these results can be used to provide a rough estimate of the accuracy
of each method. 28 mutants exhibited UV sensitive phenotypes in both the literature and
at least one of the large scale studies (this study and Birrell, et al.) and are therefore
candidates for true positive UV sensitive strains that are correct in the homozygous
diploid deletion collection. Based on this assumption and the results in Supplementary
Tables 1 and 2, we estimate false negative rates of 43% for our conservative set, 18% for
our relaxed set, and 14% for the Birrell set.


Additional UV mutants identified in this study

    Our conservative criterion also identified 14 mutants (Supplemental Table 4) with
UV sensitive phenotypes that had not been reported by the Birrell, et al. study. To
confirm the results of the high throughput assay, we tested the UV sensitivity of each
strain individually. The results (Supplemental Figure 3) demonstrate that with the
exception of one strain (cdc40) which was too sick to permit accurate measurement, all
strains displayed a detectable UV sensitive phenotype. More specifically, four of the
strains tested (cnm67, hpr1, mrlp3, and pop2_1) displayed weak, but detectable UV
sensitivity, with differences relative to the wild-type strain visible after 1 day of growth.
The remaining strains all displayed strong UV sensitivity, with differences relative to the
wild-type strain visible even after 2 days of growth. Finally, tested whether these new UV
sensitive strains contained the correct gene deletion by PCR (http://www-
sequence.stanford.edu/group/yeast_deletion_project/confirmation.html). The results
(Dutta, Dudley, and Church, unpublished data) demonstrated that all 14 mutants, except
mrpl3, contained the correct gene deletion.



        YDR364C        CDC40                          YNR052C           POP2
        YNL225C        CNM67                          YNL139C           RLR1
        YPR135W        CTF4                           YGL070C           RPB9
        YDL013W        HEX3                           YMR190C           SGS1
        YDR138W        HPR1                           YER116C           SLX8
        YMR024W        MRPL3                          YHR100C
        YKL057C        NUP120                         YLR235C

        Supplemental Table 4. New UV sensitive mutants identified in this study.
Analysis of respiratory deficiency

   We compared results from our two respiratory conditions, YPGlycerol and
YPLactate, to the respiratory mutants identified by Steinmetz et al. (Nat. Genet.
31(4):400-404). Steinmetz et al. divided the homozygous diploid mutants screened by
competitive growth and Affymetrix barcode hybridization into 4 classes based on mutant
growth in the respiratory media compared to growth in glucose medium:


                Class I: had no respiratory phenotype (3,928)
                Class II: grew better in the respiratory media (276)
                Class llla: severe growth defect in the respiratory media (214)
                Class IIIb: moderate growth defect in the respiratory media (288)


   We first compared our mutants with glycerol or lactate growth defects to the
classification of the same mutants in the Steinmetz study. In this analysis we considered
both “conservative” or high confidence results that showed growth defects in both
replicates as well as “relaxed” or lower confidence results that showed growth replicates
in only one of the replicates. We also considered mutants that showed growth defects in
“both” respiratory conditions or “either” condition alone. The results (Table I), show that
~95% of the mutants identified by the relaxed (both) and conservative (either and both)
criteria were also identified as respiratory deficient (class 3) by the Steinmetz study and
can thus be considered “true positives.” Interestingly, using the relaxed (both) criteria
nearly doubles the number of mutants identified without increasing the number of false
positives.


                              1      2      3a      3b      no match       total     % resp*
Relaxed (both)                9      3     135       88         3           238       95%
Relaxed (either)             70     13     160      133         6           382       78%
Conservative (both)           6      1       83       46        0           136       95%
Conservative (either)        13      3     122        87        4           229       93%

*% resp = (3a + 3b)/ (Total – No Match)

Supplemental Table 5. Comparison of our respiratory mutants to their classifications
assigned by Steinmetz, et al.
   We also wanted to assess our coverage of the Steinmetz respiratory mutants, i.e. what
percentage of their mutants we identified in our study. The results are shown in Table II.
Consistent with the hypothesis that the competitive growth assay is a more sensitive
assay for weaker phenotypes, we identified ~2x as many mutants in the Steinmetz severe
class (3a) as in the less severe class (3b). In addition to the possible increased sensitivity
of the competitive growth method, a major difference between the experiments was the
Steinmetz group’s use of a larger number of respiratory conditions, including the use of a
respiratory uncoupling agent, which could help identify additional mutants with mild or
no growth defects in glycerol or lactate media.

                                                 3a*                   3b*
                     Relaxed (both)        135/ 212 (64%)        88/ 277 (32%)
                     Relaxed (either)      160/ 212 (75%)       133/ 277 (48%)
                     Conserv. (both)        83/ 212 (39%)         46/ 277 (17%)
                     Conserv. (either)     122/ 212 (58%)         87/ 277 (31%)


The 3a (212) and 3b (277) totals have been adjusted to only include mutants present in
our collection.

Supplemental Table 6. Coverage of the Steinmetz respiratory mutants by our dataset

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:6
posted:8/4/2012
language:English
pages:5