Designing Interpretable Convolution-Based Hybrid Networks for Genomics

Ghotra, Rohan, Lee, Nicholas Keone, Tripathy, Rohit, Koo, Peter K (July 2021) Designing Interpretable Convolution-Based Hybrid Networks for Genomics. bioRxiv. (Submitted)

[thumbnail of 2021.07.13.452181v1.full.pdf] PDF
2021.07.13.452181v1.full.pdf - Submitted Version
Available under License Creative Commons Attribution Non-commercial.

Download (348kB)
DOI: 10.1101/2021.07.13.452181


Hybrid networks that build upon convolutional layers with attention mechanisms have demon-strated improved performance relative to pure convolutional networks across many regulatory genome analysis tasks. Their inductive bias to learn long-range interactions provides an avenue to identify learned motif-motif interactions. For attention maps to be interpretable, the convolutional layer(s) must learn identifiable motifs. Here we systematically investigate the extent that architectural choices in convolution-based hybrid networks influence learned motif representations in first layer filters, as well as the reliability of their attribution maps generated by saliency analysis. We find that design principles previously identified in standard convolutional networks also generalize to hybrid networks. This work provides an avenue to narrow the spectrum of architectural choices when designing hybrid networks such that they are amenable to commonly used interpretability methods in genomics.

Item Type: Paper
Subjects: bioinformatics
bioinformatics > genomics and proteomics > design
bioinformatics > genomics and proteomics
bioinformatics > genomics and proteomics > genetics & nucleic acid processing > DNA, RNA structure, function, modification > genes, structure and function > gene network
CSHL Authors:
Communities: CSHL labs > Koo Lab
SWORD Depositor: CSHL Elements
Depositing User: CSHL Elements
Date: 13 July 2021
Date Deposited: 20 Dec 2023 20:15
Last Modified: 20 Dec 2023 20:15
Related URLs:

Actions (login required)

Administrator's edit/view item Administrator's edit/view item
CSHL HomeAbout CSHLResearchEducationNews & FeaturesCampus & Public EventsCareersGiving