Distinguishing cancer-associated missense mutations from common polymorphisms

Authors: Joshua S. Kaminker¹, Yan Zhang¹, Allison Waugh¹, Peter Haverty¹, Brock Peters², Dragan Sebisanovic², Jeremy Stinson², William F. Forrest³, J. Fernando Bazan⁴, Sekar Seshagiri², and Zemin Zhang¹

Affiliations: Departments of 1) Bioinformatics, 2) Molecular Biology, 3) Biostatistics, and 4) Protein Engineering, Genentech, Inc., South San Francisco, CA

• Supplementary Methods
• Random Forest predictor and required data set
  • logOddsScoresForGoTerms randomForestClassifier.R test.data.crossValidation.sift.go.pfam training.data.sift.go training.data.sift.go.pfam
• The kinome data set
  • tyrosineKinome.sequenceFiles.txt tyrosineKinome.snp.txt tyrosineKinome.somatic.txt
• The phosphatome data set
  • tyrosinePhosphatome.sequenceFiles.txt tyrosinePhosphatome.snp.txt tyrosinePhosphatome.somatic.txt
• The Ser/Thr kinase data set
  • serine_threonine.sequenceFiles.txt serine_threonine.snp.txt serine_threonine.somatic.txt
• List of candidate mutations based on EST screen
  • EST_based_Variants.xls proteinSequencesUsedInEstScreen.xls
• Data used to generate figure 1
  • cancer.protein.fasta.txt cancer.variants.txt common.variants.txt complex.protein.fasta.txt complex.variants.txt disease.variants.SwissProt.txt