Skip to Main content Skip to Navigation
Journal articles

Improving distance measures between genomic tracks with mutual proximity

Abstract : An increasing number of genomic tracks such as DNA methylation, histone modifications or transcriptomes are being produced to annotate genomes with functional states. The comparison of such high dimensional vectors obtained under various experimental conditions requires the use of a distance or dissimilarity measure. Pearson, Cosine and Lp-norm distances are commonly used for both count and binary vectors. In this article we highlight how enhancement methods such as the contrast increasing mutual proximity or local scaling improves common distance measures. We present a systematic approach to evaluate the performance of such enhanced distance measures in terms of separability of groups of experimental replicates to outline their effect. We show that the mutual proximity applied on the various distance measures drastically increases performance. Depending on the type of epigenetic experiment, mutual proximity coupled together with Pearson, Cosine, L1, Yule or Jaccard distances, proves to be highly efficient in discriminating epigenomic profiles.
Document type :
Journal articles
Complete list of metadata

https://hal.sorbonne-universite.fr/hal-03335513
Contributor : Hal Sorbonne Université Gestionnaire Connect in order to contact the contributor
Submitted on : Monday, September 6, 2021 - 11:51:40 AM
Last modification on : Monday, September 13, 2021 - 10:26:01 AM

File

 Restricted access
To satisfy the distribution rights of the publisher, the document is embargoed until : 2022-03-06

Please log in to resquest access to the document

Identifiers

Citation

Thomas Haschka, Jean Baptiste Morlot, Leopold Carron, Julien Mozziconacci. Improving distance measures between genomic tracks with mutual proximity. Briefings in Bioinformatics, Oxford University Press (OUP), 2021, ⟨10.1093/bib/bbab266⟩. ⟨hal-03335513⟩

Share

Metrics

Record views

118