Skip to Main content Skip to Navigation
Journal articles

Anomaly detection in genomic catalogues using unsupervised multi-view autoencoders

Abstract : Background: Accurate identification of Transcriptional Regulator binding locations is essential for analysis of genomic regions, including Cis Regulatory Elements. The customary NGS approaches, predominantly ChIP-Seq, can be obscured by data anomalies and biases which are difficult to detect without supervision. Results: Here, we develop a method to leverage the usual combinations between many experimental series to mark such atypical peaks. We use deep learning to perform a lossy compression of the genomic regions' representations with multiview convolutions. Using artificial data, we show that our method correctly identifies groups of correlating series and evaluates CRE according to group completeness. It is then applied to the ReMap database's large volume of curated ChIP-seq data. We show that peaks lacking known biological correlators are singled out and less confirmed in real data. We propose normalization approaches useful in interpreting black-box models. Conclusion: Our approach detects peaks that are less corroborated than average. It can be extended to other similar problems, and can be interpreted to identify correlation groups. It is implemented in an open-source tool called atyPeak.
Document type :
Journal articles
Complete list of metadata
Contributor : Odile Malbec Connect in order to contact the contributor
Submitted on : Tuesday, September 28, 2021 - 11:51:18 AM
Last modification on : Friday, April 1, 2022 - 3:54:00 AM
Long-term archiving on: : Wednesday, December 29, 2021 - 6:26:59 PM


Publisher files allowed on an open archive




Quentin Ferré, Jeanne Chèneby, Denis Puthier, Cécile Capponi, Benoît Ballester. Anomaly detection in genomic catalogues using unsupervised multi-view autoencoders. BMC Bioinformatics, BioMed Central, 2021, 22 (1), pp.460. ⟨10.1186/s12859-021-04359-2⟩. ⟨inserm-03356659⟩



Record views


Files downloads