Biology subjects

Boroevich, K. A.

Publications and source records attributed to Boroevich, K. A..

4 recordsLinked to original sources

bioRxiv2026

vDeepInsight: an injective three-dimensional voxel carrier for tabular-feature neighborhood learning

DeepInsight-style methods make tabular feature relationships accessible to convolutional networks by placing each feature at a fixed position on an image carrier. An open design question is how the carrier geometry should be constructed when feature neighborhoods themselves carry part of the signal. We introduce vDeepInsight, an injective three-dimensional (3D) voxel carrier that preserves feature neighborhoods more faithfully than matched two-dimensional (2D) carriers while keeping a one-to-one mapping from each feature to a single voxel. Genes are embedded with t-SNE or UMAP, assigned one-to-one to a sparse voxel grid by linear-sum assignment, and processed by a submanifold sparse 3D convolutional network. We evaluate the carrier on gene expression through four linked analyses. First, representation-quality metrics show that 3D layouts reduce gene-neighborhood distortion relative to matched 2D layouts before any model is trained. Second, controlled synthetic tasks show that a sparse 3D convolution can exploit this preserved locality, but only when the supervised signal is constructed to depend on co-located genes and the receptive field spans adjacent voxels. Third, on real omics tasks the 3D carrier matches or exceeds tuned tabular baselines and consistently exceeds matched 2D carriers; the margin is small on marker-type classification, where individual genes already carry much of the label (tissue, lineage and cancer-type classification), and larger on program-type tasks, where the target depends on coordinated, pathway-level multi-gene activity (drug-response regression, TCGA immunogenomic-context regression and mechanism-of-action classification). Fourth, because the assignment is injective, voxel attribution maps directly back to genes, enabling gene-level attribution and pathway-level functional interpretation without voxel-to-gene deconvolution. Overall, the added carrier dimension improves the fidelity of feature-neighborhood representation and translates this improvement into prediction gains that are largest when the signal is distributed across local gene programs rather than dominated by individual marker genes.

bioinformatics↗

bioRxiv2026

Dynamic multimodal survival prediction in multiple myeloma integrating gene expression, longitudinal laboratories, and treatment history

Prognostic stratification in multiple myeloma (MM) relies on staging systems that assign patients to fixed categories at diagnosis and discard the temporal information that accumulates during treatment. We developed a dynamic multimodal framework that predicts residual overall survival using observation windows ranging from 1 to 18 months post-diagnosis. The model integrates DeepInsight-transformed gene expression representation, longitudinal laboratory measurement trajectories across 10 analytes, and treatment history for three drug classes through an adaptive fusion mechanism that accounts for missing clinical observations. On the MMRF CoMMpass cohort (n = 752), five-fold cross-validation yielded a concordance index (C-index) of 0.773 {+/-} 0.024 and a time-dependent AUC at a 1-year prediction horizon (tdAUC1yr) of 0.789 {+/-} 0.021, outperforming all evaluated baseline methods including DeepSurv (0.633 {+/-} 0.095) and random survival forests (0.636 {+/-} 0.024) on matched cross-validation splits. Modality ablation identified longitudinal laboratory measurements as the strongest individual contributor (C-index 0.693); the DeepInsight spatial encoding of gene expression yielded higher discrimination than a multilayer perceptron (MLP) baseline operating on the same features (0.624 vs. 0.596). Kaplan-Meier analysis showed significant prognostic group separation at all primary landmarks (log-rank p < 0.001; hazard ratios 3.46-3.93). A distilled student model retaining only the DeepInsight representation and five baseline clinical features achieved C-index 0.672 and tdAUC1yr 0.740 on an independent microarray cohort (GSE24080, n = 507) without retraining. Interpretability analysis identified prognostic associations consistent with established myeloma biology, including ubiquitin-proteasome pathway genes, endoplasmic reticulum stress markers, and Interferon Alpha Response pathway enrichment.

bioinformatics↗

bioRxiv2025

scHDeepInsight: A Hierarchical Deep Learning Framework for Precise Immune Cell Annotation in Single-Cell RNA-seq Data

Immune cell classification from single-cell RNA sequencing (scRNA-seq) presents significant challenges due to complex hierarchical relationships among cell types. We introduce scHDeepInsight, a deep learning framework that extends our previous scDeepInsight model by integrating a biologically-informed classification architecture with an adaptive hierarchical focal loss. The framework leverages our established method of transforming gene expression data into two-dimensional structured images for use with convolutional neural networks by effectively capturing both global and fine-grained transcriptomic features, overcoming the limitations of flat classification approaches that ignore hierarchical relationships between cell types. scHDeepInsight dynamically adjusts loss contributions to balance performance across the hierarchy levels. It also employs STACAS batch correction, robust random masking, and SHAP-based interpretability to enhance prediction accuracy and biological insight. Comprehensive benchmarking across seven diverse tissue datasets shows scHDeepInsight achieves an average accuracy of 93.2%, representing a 5.1 percentage point improvement over current state-of-the-art methods. The model successfully distinguishes 50 distinct immune cell subtypes with high accuracy, demonstrating proficiency for identifying rare and closely related cell subtypes. These advantages make scHDeepInsight a robust tool for high-resolution immune cell subtype characterization, well suited for detailed immune profiling in immunological studies.

bioinformatics↗

bioRxiv2017

Discovery and characterization of coding and non-coding driver mutations in more than 2,500 whole cancer genomes

Discovery of cancer drivers has traditionally focused on the identification of protein-coding genes. Here we present a comprehensive analysis of putative cancer driver mutations in both protein-coding and non-coding genomic regions across >2,500 whole cancer genomes from the Pan-Cancer Analysis of Whole Genomes (PCAWG) Consortium. We developed a statistically rigorous strategy for combining significance levels from multiple driver discovery methods and demonstrate that the integrated results overcome limitations of individual methods. We combined this strategy with careful filtering and applied it to protein-coding genes, promoters, untranslated regions (UTRs), distal enhancers and non-coding RNAs. These analyses redefine the landscape of non-coding driver mutations in cancer genomes, confirming a few previously reported elements and raising doubts about others, while identifying novel candidate elements across 27 cancer types. Novel recurrent events were found in the promoters or 5UTRs of TP53, RFTN1, RNF34, and MTG2, in the 3UTRs of NFKBIZ and TOB1, and in the non-coding RNA RMRP. We provide evidence that the previously reported non-coding RNAs NEAT1 and MALAT1 may be subject to a localized mutational process. Perhaps the most striking finding is the relative paucity of point mutations driving cancer in non-coding genes and regulatory elements. Though we have limited power to discover infrequent non-coding drivers in individual cohorts, combined analysis of promoters of known cancer genes show little excess of mutations beyond TERT.

genomics↗