Direct allelic variation scanning of the yeast genome

Science

Volumn 281, Issue 5380, 1998, Pages 1194-1197

  Winzeler, Elizabeth A ^a   Richards, Dan R ^a   Conway, Andrew R ^a   Goldstein, Alan L ^b   Kalman, Sue ^a   McCullough, Michael J ^a   McCusker, John H ^b   Stevens, David A ^a   Wodicka, Lisa ^c   Lockhart, David J ^c   Davis, Ronald W ^a  

^a STANFORD UNIVERSITY SCHOOL OF MEDICINE   (United States)

^b DUKE UNIVERSITY MEDICAL CENTER   (United States)

^c Thermo Fisher Scientific   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

FUNGAL DNA; OLIGONUCLEOTIDE;

ALLELE; ARTICLE; DNA HYBRIDIZATION; FUNGAL GENETICS; GENE LOCUS; GENETIC MARKER; GENETIC VARIABILITY; GENOME; MULTIDRUG RESISTANCE; NONHUMAN; PRIORITY JOURNAL; STRAIN DIFFERENCE;

EID: 0032555671     PISSN: 00368075     EISSN: None     Source Type: Journal    
DOI: 10.1126/science.281.5380.1194     Document Type: Article

References (33)

1. A. J. Schafer and J. R. Hawkins, Nature Biotechnol. 16, 33 (1998).
2. F. S. Collins, M. S. Guyer, A. Charkravarti, Science 278, 1580 (1997).
3. L. Wodicka, H. Dong, M. Mittmann, M.-H. Ho, D. J. Lockhart, Nature Biotechnol. 15, 1359 (1997).
4. 7 copies of the specific oligonucleotide probe covalently attached to a glass surface. Excluding the rDNA and CUP1 repeats, the largest gap is 41,325 bases wide at position 510,000 on chr XII.
5. 7 copies of the specific oligonucleotide probe covalently attached to a glass surface. Excluding the rDNA and CUP1 repeats, the largest gap is 41,325 bases wide at position 510,000 on chr XII.
6. B. J. Conner et al., Proc. Natl. Acad. Sci. U.S.A. 80, 278 (1983).
7. M. Chee et al., Science 274, 610 (1996).
8. S. F. Nelson et al., Nature Genet. 4, 11 (1993).
9. The completed S. cerevisiae genome sequence is from strain S288c, and 88% of the S288c genome is derived from EM93, which was isolated from a rotting fig near Merced, California, in 1938 [R. K. Mortimer and J. R. Johnston, Genetics 113, 35 (1986)]. S96 is isogenic with S288c but is unable to undergo mating type switching (ho), is able to mate with YJM789, and contains a lesion in the lys5 gene that can be easily scored in crosses. YJM789 is isogenic with YJM145, a segregant of a clinical isolate of S. cerevisiae (27). YJM145 has been characterized genetically, and the ultimate source of its parent (human lung) differs substantially from that of S288c in that the strains were isolated from different environments, at different times, and in different geographic locations. Theoretically, any two yeast strains could be used.
10. A library of YJM789 genomic DNA was constructed in an M13 sequencing vector. The sequence was determined for 696 clones as previously described [F. S. Dietrich et al., Nature 387, 78 (1997)]. The sequences were called by means of the phred base caller software (see chimera.biotech.washington.edu/UWGC/tools/phred.htm), which produces a quality measurement for each base [-10 x log(10) (probability of an error)]. A total of 122,258 bases were sequenced with greater than 99% confidence by this quality measurement. The YJM789 sequences were compared with S288c sequence with the cross_match program (see chimera.biotech.washington.edu/UWGC/tools/phrap.htm). Discrepancies between the YJM789 and S288c sequences were then classified by quality and assigned into coding and noncoding regions with the phred base caller. In most cases, because only a single trace was available and no alignments were performed, regions of the traces that did not show high quality were excluded from the analysis. When a high-quality sequence (>99.7% accuracy) was used, 466 cases of allelic variation were observed with a frequency of about one every 160 bases.
11. Of the 466 cases of allelic variation in sequences with greater than 99.7% accuracy, 288 were from coding regions (61%). Of the estimated 13.2 Mb, 8.637 Mb (65%) of the yeast genome is annotated as coding sequence by Saccharomyces Genome Database.
12. 2 for 5 min at 37°C. After heat inactivation of the DNase I, the DNA fragments were end-labeled in the same buffer by the addition of 25 U of terminal transferase (Boehringer Mannheim) and 1 nmol biotin-N6-dideoxyadenosine triphosphate (NEN) for 1 hour at 37°C. The entire sample was hybridized to the array in a 200-μl volume as previously described (3).
13. Grids were aligned to the scanned images by the known feature dimensions of the array. The hybridization intensities for each of the elements in the grid were determined by the seventy-fifth percentile method in the Affymetrix GeneChip software package.
14. An adjusted array hybridization intensity value (l) was determined for each hybridization (20 altogether) as the mean of the log(PM) signals of all features that showed minimal variation across all hybridizations (the nonmarkers, determined recursively as described below). Then, for each feature on the array, a linear regression of log[perfect match (PM)] on / for all hybridizations was determined by the least squares method, first under the null hypothesis that the S96 and YJM789 samples had the same response and then under the alternative hypothesis that the S96 samples had a greater signal than the YJM789 samples. The models were compared with the F test, and the same signal model was rejected in favor of a marker with 99% confidence. This software is available upon request to D. Richards.
15. Gaps were often found near regions with low probe coverage, for example, near repeated elements in the genome or regions of low open reading frame (ORF) density. However, in some cases, probe coverage was adequate, suggesting that the gap might be due to a high amount of sequence conservation or to the region having a recent common origin for the two strains.
16. The p is computed as P(S96)/[P(S96)+P(YJM789)], where P(X) is the probability (from the t distribution) that a marker has genotype X, based on the observed (PM) hybridization signal of the feature and the expected signal (given the array hybridization intensity) and the estimated variance from the regression for the marker.
17. J. M. Cherry et al., Nature 387, 67 (1997).
18. T. Petes, R. Malone, L. Symington, in The Molecular and Cellular Biology of the Yeast Saccharomyces, J. Broach, J. Pringle, E. Jones, Eds. (Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1991), vol. I, pp. 407-521.
19. B. Chattoo et al., Genetics 93, 51 (1979).
20. E. Alani, L. Cao, N. Kleckner, ibid. 116, 541 (1987).
21. Yeast strains were routinely grown in yeast extract, peptone, and dextrose (YEPD) medium; sporulation medium and defined medium for scoring auxotrophs were prepared as previously described [F. Sherman, G. Fink, C. Lawrence, Methods in Yeast Genetics: Laboratory Manual (Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1974)]. Segregants were complementation tested to distinguish lys2 from lys5. Cycloheximide sensitivity was scored by inability to grow on YEPD plates containing cycloheximide (0.5 μg/ml). All segregants from the 99 tetrads were phenotyped for lys2, lys5, and cyh. lys5 and cyh segregated 2:2 for 99 tetrads, and lys2 segregated 2:2 for 98 of the 99. Selected segregants were scored for MAT and ho. The ho and ho::hisG were distinguished by checking the size of PCR products on a gel, and MAT was determined by mating and complementation.
22. The loci could have been mapped with any segregant as long as the genotype was known; however, segregants with similar genotypes were chosen to simplify the analysis.
23. N. In general, up to 13 segregants (or more if the trait is non-Mendelian) may need to be examined to have a 95% probability of identifying a single region as responsible for a trait.
24. The breakpoints were recursively added to each chromosome on the basis of the p values. The probabilities of breakpoints at every pair of markers were tested against the probability of no breakpoint. The breakpoints that maximized this likelihood were accepted if the logarithmic likelihood ratio was greater than 30. This procedure was repeated for each new subinterval created by a breakpoint to 500-bp resolution.
25. M. Boehnke, Am. J. Hum. Genet. 55, 379 (1994).
26. E. Balzi, M. Wang, S. Leterme, L. Van Dyck, A. Goffeau, J. Biol. Chem. 269, 2206 (1994).
27. E. A. Winzeler et al., unpublished data.
28. J. H. McCusker, K. V. Clemons, D. A. Stevens, R. W. Davis, Genetics 136, 1261 (1994).
29. D. G. Wang et al., Science 280, 1077 (1998).
30. P. A. Underhill et al., Genome Res. 7, 996 (1997).
31. M. Orita, H. Iwahana, H. Kanazawa, K. Hayashi, T. Sekiya, Proc. Natl. Acad. Sci. U.S.A. 86, 2766 (1989).
32. D. J. Lockhart et al., Nature Biotechnol. 14, 1675 (1996).
33. We thank N. Risch and D. Siegmund for helpful advice. E.A.W is supported by the John Wasmuth Fellowship in Genomic Analysis (HG00185-01). D.R.R is a Howard Hughes Medical Institute predoctoral fellow. M.J.M. is supported by a Commonwealth AIDS Research Grants Committee of Australia postdoctoral overseas fellowship. Funding by NIH grant 1R01 HG01633.