APNet, an Explainable Sparse Deep Learning Model to Discover Differentially Active Drivers of Severe COVID-19

Overview

Journal Bioinformatics

Publisher Oxford University Press

Date 2025 Feb 8

PMID 39921901

Authors

George I Gavriilidis

Vasileios Vasileiou

Stella Dimitsaki

Georgios Karakatsoulis

Antonis Giannakakis

Georgios A Pavlopoulos

Fotis Psomopoulos

Affiliations

Soon will be listed here.

Abstract

Motivation: Computational analyses of bulk and single-cell omics provide translational insights into complex diseases, such as COVID-19, by revealing molecules, cellular phenotypes, and signalling patterns that contribute to unfavourable clinical outcomes. Current in silico approaches dovetail differential abundance, biostatistics, and machine learning, but often overlook nonlinear proteomic dynamics, like post-translational modifications, and provide limited biological interpretability beyond feature ranking.

Results: We introduce APNet, a novel computational pipeline that combines differential activity analysis based on SJARACNe co-expression networks with PASNet, a biologically informed sparse deep learning model, to perform explainable predictions for COVID-19 severity. The APNet driver-pathway network ingests SJARACNe co-regulation and classification weights to aid result interpretation and hypothesis generation. APNet outperforms alternative models in patient classification across three COVID-19 proteomic datasets, identifying predictive drivers and pathways, including some confirmed in single-cell omics and highlighting under-explored biomarker circuitries in COVID-19.

Availability And Implementation: APNet's R, Python scripts, and Cytoscape methodologies are available at https://github.com/BiodataAnalysisGroup/APNet.

References

Davies D, Bouldin D . A cluster separation measure. IEEE Trans Pattern Anal Mach Intell. 2011; 1(2):224-7. View

Walsh I, Fishman D, Garcia-Gasulla D, Titma T, Pollastri G, Harrow J . DOME: recommendations for supervised machine learning validation in biology. Nat Methods. 2021; 18(10):1122-1127. DOI: 10.1038/s41592-021-01205-4. View

Sapoval N, Aghazadeh A, Nute M, Antunes D, Balaji A, Baraniuk R . Current progress and open challenges for applying deep learning across the biosciences. Nat Commun. 2022; 13(1):1728. PMC: 8976012. DOI: 10.1038/s41467-022-29268-7. View

Filbin M, Mehta A, Schneider A, Kays K, Guess J, Gentili M . Longitudinal proteomic analysis of severe COVID-19 reveals survival-associated signatures, tissue-specific cell death, and cell-cell interactions. Cell Rep Med. 2021; 2(5):100287. PMC: 8091031. DOI: 10.1016/j.xcrm.2021.100287. View

Zhong W, Edfors F, Gummesson A, Bergstrom G, Fagerberg L, Uhlen M . Next generation plasma proteome profiling to monitor health and disease. Nat Commun. 2021; 12(1):2493. PMC: 8093230. DOI: 10.1038/s41467-021-22767-z. View

Mothes R, Pascual-Reguant A, Koehler R, Liebeskind J, Liebheit A, Bauherr S . Distinct tissue niches direct lung immunopathology via CCL18 and CCL21 in severe COVID-19. Nat Commun. 2023; 14(1):791. PMC: 9922044. DOI: 10.1038/s41467-023-36333-2. View

Wysocka M, Wysocki O, Zufferey M, Landers D, Freitas A . A systematic review of biologically-informed deep learning models for cancer: fundamental trends for encoding and interpreting oncology data. BMC Bioinformatics. 2023; 24(1):198. PMC: 10186658. DOI: 10.1186/s12859-023-05262-8. View

Penrice-Randal R, Dong X, Shapanis A, Gardner A, Harding N, Legebeke J . Blood gene expression predicts intensive care unit admission in hospitalised patients with COVID-19. Front Immunol. 2022; 13:988685. PMC: 9530807. DOI: 10.3389/fimmu.2022.988685. View

Byeon S, Madugundu A, Garapati K, Ramarajan M, Saraswat M, Kumar-M P . Development of a multiomics model for identification of predictive biomarkers for COVID-19 severity: a retrospective cohort study. Lancet Digit Health. 2022; 4(9):e632-e645. PMC: 9273185. DOI: 10.1016/S2589-7500(22)00112-1. View

10.

Hao J, Kim Y, Mallavarapu T, Oh J, Kang M . Interpretable deep neural network for cancer survival analysis by integrating genomic and clinical data. BMC Med Genomics. 2019; 12(Suppl 10):189. PMC: 6927105. DOI: 10.1186/s12920-019-0624-2. View

11.

. A blood atlas of COVID-19 defines hallmarks of disease severity and specificity. Cell. 2022; 185(5):916-938.e58. PMC: 8776501. DOI: 10.1016/j.cell.2022.01.012. View

12.

Babacic H, Christ W, Araujo J, Mermelekas G, Sharma N, Tynell J . Comprehensive proteomics and meta-analysis of COVID-19 host response. Nat Commun. 2023; 14(1):5921. PMC: 10516886. DOI: 10.1038/s41467-023-41159-z. View

13.

Narayanan S, Jamison Jr D, Guarnieri J, Zaksas V, Topper M, Koutnik A . A comprehensive SARS-CoV-2 and COVID-19 review, Part 2: host extracellular to systemic effects of SARS-CoV-2 infection. Eur J Hum Genet. 2023; 32(1):10-20. PMC: 10772081. DOI: 10.1038/s41431-023-01462-1. View

14.

Santorsola M, Lescai F . The promise of explainable deep learning for omics data analysis: Adding new discovery tools to AI. N Biotechnol. 2023; 77:1-11. DOI: 10.1016/j.nbt.2023.06.002. View

15.

Feyaerts D, Hedou J, Gillard J, Chen H, Tsai E, Peterson L . Integrated plasma proteomic and single-cell immune signaling network signatures demarcate mild, moderate, and severe COVID-19. Cell Rep Med. 2022; 3(7):100680. PMC: 9238057. DOI: 10.1016/j.xcrm.2022.100680. View

16.

Khatamian A, Paull E, Califano A, Yu J . SJARACNe: a scalable software tool for gene network reverse engineering from big data. Bioinformatics. 2018; 35(12):2165-2166. PMC: 6581437. DOI: 10.1093/bioinformatics/bty907. View

17.

Wik L, Nordberg N, Broberg J, Bjorkesten J, Assarsson E, Henriksson S . Proximity Extension Assay in Combination with Next-Generation Sequencing for High-throughput Proteome-wide Analysis. Mol Cell Proteomics. 2021; 20:100168. PMC: 8633680. DOI: 10.1016/j.mcpro.2021.100168. View

18.

Hao J, Kim Y, Kim T, Kang M . PASNet: pathway-associated sparse deep neural network for prognosis prediction from high-throughput data. BMC Bioinformatics. 2018; 19(1):510. PMC: 6296065. DOI: 10.1186/s12859-018-2500-z. View

19.

Lee J, Park S, Jeong H, Ahn J, Choi S, Lee H . Immunophenotyping of COVID-19 and influenza highlights the role of type I interferons in development of severe COVID-19. Sci Immunol. 2020; 5(49). PMC: 7402635. DOI: 10.1126/sciimmunol.abd1554. View

20.

Margolin A, Nemenman I, Basso K, Wiggins C, Stolovitzky G, Dalla Favera R . ARACNE: an algorithm for the reconstruction of gene regulatory networks in a mammalian cellular context. BMC Bioinformatics. 2006; 7 Suppl 1:S7. PMC: 1810318. DOI: 10.1186/1471-2105-7-S1-S7. View