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

Abstract

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 GI50 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 GI50 thresholds. Ensembles of gene signatures at different thresholds are combined to reduce the dependence on specific GI50 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.

Single Spheroid Metabolomics: Optimizing Sample Preparation of Three-Dimensional Multicellular Tumor Spheroids

“…Finally, the MetaboAnalyst 4.0 Pathway Analysis module [46] revealed another interesting pathway to be further investigated, namely, the "pantothenate and CoA biosynthesis" pathway (through uracil downregulation). A recent study [29] addressed signature genes for patients responding to oxaliplatin therapy by machine learning. Among the most accurate signature genes for oxaliplatin treatment was PANK3 which encodes for pantothenate kinase, a key regulatory enzyme in the biosynthesis of coenzyme A (CoA).…”

“…Metal-based drugs are a prime example since a clear cut mechanism remains to be elucidated despite extensive clinical use and fundamental research [26,27]. In fact, how the drugs exert their cytotoxicity differs even for the three clinically approved platinum(II) drugs [28][29][30]. Although massive research efforts resulted in a plethora of promising candidate drugs, the failure rate upon translation into clinics was/is extremely high for metal-based anticancer drugs [31].…”

Estimating partial-body ionizing radiation exposure by automated cytogenetic biodosimetry

“…Partial body exposures could also be quantified using next generation sequencing-based RNA-Seq data that distinguishes constitutional-from radiation-specific, alternatively spliced transcript read counts. These features could be incorporated into biochemically inspiredmachine learning-based gene expression signatures of ionizing radiation (Dorman et al 2016;Macaeva et al 2016;Mucaki et al 2016;Zhao et al 2018;Mucaki et al 2019;Mucaki et al 2020).…”

Prediction of Caesarean Delivery

“…This study examined the responses of chemotherapy agents cisplatin, carboplatin and oxaliplatin with certain gene signatures [61]. This group developed a machine-learning based prediction model, which aimed to predict the effectiveness of the agents above to certain gene signatures.…”