dna2vec: Consistent vector representations of variable-length k-mers

Patrick Ng
1/23/2017
q-bio.QMcs.CLcs.LG

Official code: pnpnpn/dna2vec (193 stars, MIT, last commit 2022-06-21).

Abstract

One of the ubiquitous representation of long DNA sequence is dividing it into shorter k-mer components. Unfortunately, the straightforward vector encoding of k-mer as a one-hot vector is vulnerable to the curse of dimensionality. Worse yet, the distance between any pair of one-hot vectors is equidistant. This is particularly problematic when applying the latest machine learning algorithms to solve problems in biological sequence analysis. In this paper, we propose a novel method to train distributed representations of variable-length k-mers. Our method is based on the popular word embedding model word2vec, which is trained on a shallow two-layer neural network. Our experiments provide evidence that the summing of dna2vec vectors is akin to nucleotides concatenation. We also demonstrate that there is correlation between Needleman-Wunsch similarity score and cosine similarity of dna2vec vectors.

View on arXivView PDF

Code Implementations(1)

pnpnpn/dna2vecOfficial100%
MIT

dna2vec: Consistent vector representations of variable-length k-mers

19361PythonMar 6, 20174 years agoMIT
bioinformaticscomputational-biologyembeddingsmachine-learningml+5 more

Similar Papers

Learning Genomic Sequence Representations using Graph Neural Networks over De Bruijn Graphs

80% match1 repo72023

Dy-mer: An Explainable DNA Sequence Representation Scheme using Dictionary Learning

80% match2024

Learning from String Sequences

80% match2024

HyCoSeq: Contextual Hyperbolic Representation Learning for Genomic Sequences

79% match2026

Vector Embeddings by Sequence Similarity and Context for Improved Compression, Similarity Search, Clustering, Organization, and Manipulation of cDNA Libraries

79% match2023

DNAZEN: Enhanced Gene Sequence Representations via Mixed Granularities of Coding Units

78% match1 repo32025

Cite this paper

@article{ng2017dnavec,
  title  = {dna2vec: Consistent vector representations of variable-length k-mers},
  author = {Patrick Ng},
  year   = {2017},
  eprint = {1701.06279},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1701.06279}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1701.06279.svg)](https://codeofpaper.com/paper/1701.06279)

Discussion