Search

SummaryIdentification of gene expression state of a cancer patient from routine pathology imaging and characterization of its phenotypic effects have significant clinical and therapeutic implications. However, prediction of expression of individual genes from whole slide images (WSIs) is challenging due to co-dependent or correlated expression of multiple genes. Here, we use a purely data-driven approach to first identify groups of genes with co-dependent expression and then predict their status from (WSIs) using a bespoke graph neural network. These gene groups allow us to capture the gene expression state of a patient with a small number of binary variables that are biologically meaningful and carry histopathological insights for clinically and therapeutic use cases. Prediction of gene expression state based on these gene groups allows associating histological phenotypes (cellular composition, mitotic counts, grading, etc.) with underlying gene expression patterns and opens avenues for gaining significant biological insights from routine pathology imaging directly.Graphical AbstractHighlightsData-driven discovery of co-expressing gene groups in breast canerHistological imaging based prediction of gene groups via deep learningIdentification of phenotypic correlates of gene-expression in histological imagingClinical and therapeutic impact of gene groups and their visual patterns identified

Motivation Machine-learning-based prediction of compound–protein interactions (CPIs) is important for drug design, screening and repurposing. Despite numerous recent publication with increasing methodological sophistication claiming consistent improvements in predictive accuracy, we have observed a number of fundamental issues in experiment design that produce overoptimistic estimates of model performance. Results We systematically analyze the impact of several factors affecting generalization performance of CPI predictors that are overlooked in existing work: (i) similarity between training and test examples in cross-validation; (ii) synthesizing negative examples in absence of experimentally verified negative examples and (iii) alignment of evaluation protocol and performance metrics with real-world use of CPI predictors in screening large compound libraries. Using both state-of-the-art approaches by other researchers as well as a simple kernel-based baseline, we have found that effective assessment of generalization performance of CPI predictors requires careful control over similarity between training and test examples. We show that, under stringent performance assessment protocols, a simple kernel-based approach can exceed the predictive performance of existing state-of-the-art methods. We also show that random pairing for generating synthetic negative examples for training and performance evaluation results in models with better generalization in comparison to more sophisticated strategies used in existing studies. Our analyses indicate that using proposed experiment design strategies can offer significant improvements for CPI prediction leading to effective target compound screening for drug repurposing and discovery of putative chemical ligands of SARS-CoV-2-Spike and Human-ACE2 proteins. Availability and implementation Code and supplementary material available at https://github.com/adibayaseen/HKRCPI. Supplementary information Supplementary data are available at Bioinformatics online.

Runxuan Zhang: et al., [Background] Accurate and comprehensive annotation of transcript sequences is essential for transcript quantification and differential gene and transcript expression analysis. Single-molecule long-read sequencing technologies provide improved integrity of transcript structures including alternative splicing, and transcription start and polyadenylation sites. However, accuracy is significantly affected by sequencing errors, mRNA degradation, or incomplete cDNA synthesis., [Results] We present a new and comprehensive Arabidopsis thaliana Reference Transcript Dataset 3 (AtRTD3). AtRTD3 contains over 169,000 transcripts—twice that of the best current Arabidopsis transcriptome and including over 1500 novel genes. Seventy-eight percent of transcripts are from Iso-seq with accurately defined splice junctions and transcription start and end sites. We develop novel methods to determine splice junctions and transcription start and end sites accurately. Mismatch profiles around splice junctions provide a powerful feature to distinguish correct splice junctions and remove false splice junctions. Stratified approaches identify high-confidence transcription start and end sites and remove fragmentary transcripts due to degradation. AtRTD3 is a major improvement over existing transcriptomes as demonstrated by analysis of an Arabidopsis cold response RNA-seq time-series. AtRTD3 provides higher resolution of transcript expression profiling and identifies cold-induced differential transcription start and polyadenylation site usage., [Conclusions] AtRTD3 is the most comprehensive Arabidopsis transcriptome currently. It improves the precision of differential gene and transcript expression, differential alternative splicing, and transcription start/end site usage analysis from RNA-seq data. The novel methods for identifying accurate splice junctions and transcription start/end sites are widely applicable and will improve single-molecule sequencing analysis from any species., This work was jointly supported by funding from the Biotechnology and Biological Sciences Research Council (BBSRC) BB/P009751/1 to JB; BB/R014582/1 to RW and RZ; BB/S020160/1 to RZ; BB/S004610/1 (16 ERA-CAPS BARN) to RW; the Scottish Government Rural and Environment Science and Analytical Services division (RESAS) [to RZ, RW, and JB]; the National Science Foundation (MCB-2014408) and the National Institute of Health (NIH) (GM-114297) to E.H.; S. H. was supported by funding to K.D. from the University of York; the Austrian Science Fund (FWF) SFB F43 to AB and MJ and [P26333] to MK; The French Agence Nationale de la Recherche grant ANR-16-CE12-0032 to MC; the Japan Science and Technology Agency (JST), the Core Research for Evolutionary Science and Technology (CREST; Grant Number JPMJCR13B4) to M.S.; the National Science Foundation (Grant No. DBI1949036 to A.b.H and A.S.N.R, and Grant No. MCB 2014542 to E.H. and A.S.N.R.); and the DOE Office of Science, Office of Biological and Environmental Research (Grant No. DE-SC0010733) to A.S.N.R and A.b.H.; the Deutsche Forschungsgemeinschaft (DFG) STA653/14-1 and STA653/15-1 to DS; the National Science Foundation grant (IOS-154173) to Q.Q.L.; the German Research Foundation (DFG) WA2167/8-1 to AW and SFB1101/C03 to AW and TWK; the Research Grants Council (RGC) of Hong Kong (GRF 12103020) to LX. NSF grant IOS-1849708 and NSF EPSCoR grant 1826836 to RS; the Academia Sinica to S.-L. T.

Alternative splicing is a widespread regulatory phenomenon that enables a single gene to produce multiple transcripts. Among the different types of alternative splicing, intron retention is one of the least explored despite its high prevalence in both plants and animals. The recent discovery that the majority of splicing is co-transcriptional has led to the finding that chromatin state affects alternative splicing. Therefore it is plausible that transcription factors can regulate splicing outcomes. We provide evidence for this hypothesis by studying regions of open chromatin in retained and excised introns. Using deep learning models designed to distinguish between regions of open chromatin in retained introns and non-retained introns, we identified motifs enriched in IR events with significant hits to known human transcription factors. Our model predicts that the majority of transcription factors that affect intron retention come from the zinc finger family. We demonstrate the validity of these predictions using ChIP-seq data for multiple zinc finger transcription factors and find strong over-representation for their peaks in intron retention events.AvailabilitySource code available at https://github.com/fahadahaf/chromir