Interpretation, Stratification and Evidence for Sequence Variants Affecting mRNA Splicing in Complete Human Genome Sequences

Information theory-based methods have been shown to be sensitive and specific for pre- dicting and quantifying the effects of non-coding mutations in Mendelian diseases. We present the Shannon pipeline software for genome-scale mutation analysis and provide evidence that the soft- ware predicts variants affecting mRNA splicing. Individual information contents （in bits） of refer- ence and variant splice sites are compared and significant differences are annotated and prioritized. The software has been implemented for CLC-Bio Genomics platform. Annotation indicates the context of novel mutations as well as common and rare SNPs with splicing effects. Potential natural and cryptic mRNA splicing variants are identified, and null mutations are distinguished from leaky mutations. Mutations and rare SNPs were predicted in genomes of three cancer cell lines （U2OS, U251 and A431）, which were supported by expression analyses. After filtering, tractable numbers of potentially deleterious variants are predicted by the software, suitable for further laboratory investigation. In these cell lines, novel functional variants comprised 6-17 inactivating mutations, 1 5 leaky mutations and 6-13 cryptic splicing mutations. Predicted effects were validated by RNA-seq analysis of the three aforementioned cancer cell lines, and expression microarray analysis of SNPs in HapMap cell lines.

Predicting responses to platin chemotherapy agents with biochemically-inspired machine learning

The selection of effective genes that accurately predict chemotherapy responses might improve cancer outcomes.We compare optimized gene signatures for cisplatin,carboplatin,and oxaliplatin responses in the same cell lines and validate each signature using data from patients with cancer.Supervised support vector machine learning is used to derive gene sets whose expression is related to the cell line GI 50 values by backwards feature selection with cross-validation.Specific genes and functional pathways distinguishing sensitive from resistant cell lines are identified by contrasting signatures obtained at extreme and median GI 50 thresholds.Ensembles of gene signatures at different thresholds are combined to reduce the dependence on specific GI 50 values for predicting drug responses.The most accurate gene signatures for each platin are:cisplatin:BARD1,BCL2,BCL2L1,CDKN2C,FAAP24,FEN1,MAP3K1,MAPK13,MAPK3,NFKB1,NFKB2,SLC22A5,SLC31A2,TLR4,and TWIST1;carboplatin:AKT1,EIF3K,ERCC1,GNGT1,GSR,MTHFR,NEDD4L,NLRP1,NRAS,RAF1,SGK1,TIGD1,TP53,VEGFB,and VEGFC;and oxaliplatin:BRAF,FCGR2A,IGF1,MSH2,NAGK,NFE2L2,NQO1,PANK3,SLC47A1,SLCO1B1,and UGT1A1.Data from The Cancer Genome Atlas(TCGA)patients with bladder,ovarian,and colorectal cancer were used to test the cisplatin,carboplatin,and oxaliplatin signatures,resulting in 71.0%,60.2%,and 54.5%accuracies in predicting disease recurrence and 59%,61%,and 72%accuracies in predicting remission,respectively.One cisplatin signature predicted 100%of recurrence in non-smoking patients with bladder cancer(57%disease-free;N=19),and 79%recurrence in smokers(62%disease-free;N=35).This approach should be adaptable to other studies of chemotherapy responses,regardless of the drug or cancer types.