Rui Jiang | bioRxiv Science

bioRxiv2020

DeepCDR: a hybrid graph convolutional network for predicting cancer drug response

Motivation Accurate prediction of cancer drug response (CDR) is challenging due to the uncertainty of drug efficacy and heterogeneity of cancer patients. Strong evidences have implicated the high dependence of CDR on tumor genomic and transcriptomic profiles of individual patients. Precise identification of CDR is crucial in both guiding anti-cancer drug design and understanding cancer biology.Results In this study, we present DeepCDR which integrates multi-omics profiles of cancer cells and explores intrinsic chemical structures of drugs for predicting cancer drug response. Specifically, DeepCDR is a hybrid graph convolutional network consisting of a uniform graph convolutional network (UGCN) and multiple subnetworks. Unlike prior studies modeling hand-crafted features of drugs, DeepCDR automatically learns the latent representation of topological structures among atoms and bonds of drugs. Extensive experiments showed that DeepCDR outperformed state-of-the-art methods in both classification and regression settings under various data settings. We also evaluated the contribution of different types of omics profiles for assessing drug response. Furthermore, we provided an exploratory strategy for identifying potential cancer-associated genes concerning specific cancer types. Our results highlighted the predictive power of DeepCDR and its potential translational value in guiding disease-specific drug design.Availability DeepCDR is freely available at https://github.com/kimmo1019/DeepCDRContact ruijiang{at}tsinghua.edu.cn; muzhou{at}sensebrain.siteSupplementary information Supplementary data are available at Bioinformatics online.Competing Interest StatementThe authors have declared no competing interest.View Full Text

bioinformatics↗

bioRxiv2014

ProbAlign: a re-alignment method for long sequencing reads

The incorrect alignments are a severe problem in variant calling, and remain as a challenge computational issue in Bioinformatics field. Although there have been some methods utilizing the re-alignment approach to tackle the misalignments, a standalone re-alignment tool for long sequencing reads is lacking. Hence, we present a standalone tool to correct the misalignments, called ProbAlign. It can be integrated into the pipelines of not only variant calling but also other genomic applications. We demonstrate the use of re-alignment in two diverse and important genomics fields: variant calling and viral quasispecies reconstruction. First, variant calling results in the Pacific Biosciences SMRT re-sequencing data of NA12878 show that false positives can be reduced by 43.5%, and true positives can be increased by 24.8% averagely, after re-alignment. Second, results in reconstructing a 5-virus-mix show that the viral population can be completely unraveled, and also the estimation of quasispecies frequencies has been improved, after re-alignment. ProbAlign is freely available in the PyroTools toolkit (https://github.com/homopolymer/PyroTools).

Bioinformatics↗