Jonathan Z.L. Zhao's research while affiliated with University of the Western Cape and other places

Machine learning has identified genetic signatures that predict how patients will respond to three of the most widely used cancer drugs. Chemotherapy regimens are usually based on how groups of people with similar cancers respond to them, but genetic differences can render the drugs more or less effective in individual patients. Machine learning provides a way of sifting through large amounts of data to identify patterns—in this case, in gene signatures associated with cancer recurrence and remission. The authors investigated cellular responses to cisplatin, carboplatin, and oxaliplatin and identified signatures in 11–15 genes which were the most predictive for each drug. The compositions of these signatures are also tailored to how well these therapies prevent growth of cancer cells. Accuracy varied, but one cisplatin signature was able to predict all instances of disease recurrence in non-smokers with bladder cancer.

Selection of effective genes that accurately predict chemotherapy response could improve cancer outcomes. We compare optimized gene signatures for cisplatin, carboplatin, and oxaliplatin response in the same cell lines, and respectively validate each with cancer patient data. Supervised support vector machine learning was used to derive gene sets whose expression was related to 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 vs. median GI 50 thresholds. Ensembles of gene signatures at different thresholds are combined to reduce dependence on specific GI 50 values for predicting drug response. The most accurate models for each platin are: cisplatin: BARD1 , BCL2 , BCL2L1 , CDKN2C , FAAP24 , FEN1 , MAP3K1 , MAPK13 , MAPK3 , NFKB1 , NFKB2 , SLC22A5 , SLC31A2 , TLR4 , TWIST1 ; carboplatin: AKT1 , EIF3K , ERCC1 , GNGT1 , GSR , MTHFR , NEDD4L , NLRP1 , NRAS , RAF1 , SGK1 , TIGD1 , TP53 , VEGFB , VEGFC; oxaliplatin: BRAF , FCGR2A , IGF1 , MSH2 , NAGK , NFE2L2 , NQO1 , PANK3 , SLC47A1 , SLCO1B1 , UGT1A1 . TCGA bladder, ovarian and colorectal cancer patients were used to test cisplatin, carboplatin and oxaliplatin signatures (respectively), resulting in 71.0%, 60.2% and 54.5% accuracy in predicting disease recurrence and 59%, 61% and 72% accuracy in predicting remission. One cisplatin signature predicted 100% of recurrence in non-smoking bladder cancer patients (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 response, independent of drug or cancer types.

Background: Gene signatures derived from transcriptomic data using machine learning methods have shown promise for biodosimetry testing. These signatures may not be sufficiently robust for large scale testing, as their performance has not been adequately validated on external, independent datasets. The present study develops human and murine signatures with biochemically-inspired machine learning that are strictly validated using k-fold and traditional approaches. Methods: Gene Expression Omnibus (GEO) datasets of exposed human and murine lymphocytes were preprocessed via nearest neighbor imputation and expression of genes implicated in the literature to be responsive to radiation exposure (n=998) were then ranked by Minimum Redundancy Maximum Relevance (mRMR). Optimal signatures were derived by backward, complete, and forward sequential feature selection using Support Vector Machines (SVM), and validated using k-fold or traditional validation on independent datasets. Results: The best human signatures we derived exhibit k-fold validation accuracies of up to 98% ( DDB2 , PRKDC , TPP2 , PTPRE , and GADD45A ) when validated over 209 samples and traditional validation accuracies of up to 92% ( DDB2 , CD8A , TALDO1 , PCNA , EIF4G2 , LCN2 , CDKN1A , PRKCH , ENO1 , and PPM1D ) when validated over 85 samples. Some human signatures are specific enough to differentiate between chemotherapy and radiotherapy. Certain multi-class murine signatures have sufficient granularity in dose estimation to inform eligibility for cytokine therapy (assuming these signatures could be translated to humans). We compiled a list of the most frequently appearing genes in the top 20 human and mouse signatures. More frequently appearing genes among an ensemble of signatures may indicate greater impact of these genes on the performance of individual signatures. Several genes in the signatures we derived are present in previously proposed signatures. Conclusions: Gene signatures for ionizing radiation exposure derived by machine learning have low error rates in externally validated, independent datasets, and exhibit high specificity and granularity for dose estimation.

Background: Gene signatures derived from transcriptomic data using machine learning methods have shown promise for biodosimetry testing. These signatures may not be sufficiently robust for large scale testing, as their performance has not been adequately validated on external, independent datasets. The present study develops human and murine signatures with biochemically-inspired machine learning that are strictly validated using k-fold and traditional approaches. Methods: Gene Expression Omnibus (GEO) datasets of exposed human and murine lymphocytes were preprocessed via nearest neighbor imputation and expression of genes implicated in the literature to be responsive to radiation exposure (n=998) were then ranked by Minimum Redundancy Maximum Relevance (mRMR). Optimal signatures were derived by backward, complete, and forward sequential feature selection using Support Vector Machines (SVM), and validated using k-fold or traditional validation on independent datasets. Results: The best human signatures we derived exhibit k-fold validation accuracies of up to 98% (DDB2, PRKDC, TPP2, PTPRE, and GADD45A) when validated over 209 samples and traditional validation accuracies of up to 92% (DDB2, CD8A, TALDO1, PCNA, EIF4G2, LCN2, CDKN1A, PRKCH, ENO1, and PPM1D) when validated over 85 samples. Some human signatures are specific enough to differentiate between chemotherapy and radiotherapy. Certain multi-class murine signatures have sufficient granularity in dose estimation to inform eligibility for cytokine therapy (assuming these signatures could be translated to humans). We compiled a list of the most frequently appearing genes in the top 20 human and mouse signatures. More frequently appearing genes among an ensemble of signatures may indicate greater impact of these genes on the performance of individual signatures. Several genes in the signatures we derived are present in previously proposed signatures. Conclusions: Gene signatures for ionizing radiation exposure derived by machine learning have low error rates in externally validated, independent datasets, and exhibit high specificity and granularity for dose estimation.