arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1906.03087q-bio.GNcs.LG
收藏

DNA序列的无监督表示学习

Unsupervised Representation Learning of DNA Sequences

Indian Institute of Technology Guwahati(印度理工学院古瓦哈蒂分校)

Vishal Agarwal, N Jayanth Kumar Reddy, Ashish Anand

摘要:

最近,几种深度学习模型已被用于基于DNA序列的分类任务。通常,这样的任务需要输入长且可变长度的DNA序列。在这项工作中,我们使用一个序列到序列的自动编码器模型,学习一个固定尺寸的长和可变长度的DNA序列在一个无监督的方式的潜在表示。我们定量和定性评估的剪接位点分类的监督任务的学习潜在的代表性。定量评价是在两种不同的设置下进行的。我们的实验表明,这些表示可以被用作功能或先验密切相关的任务,如剪接位点分类。此外,在我们的定性分析中,我们使用模型归因技术集成的特征来推断影响分类准确性的重要序列特征。我们发现,所确定的剪接签名与现有知识相似。

英文摘要:

Recently several deep learning models have been used for DNA sequence based classification tasks. Often such tasks require long and variable length DNA sequences in the input. In this work, we use a sequence-to-sequence autoencoder model to learn a latent representation of a fixed dimension for long and variable length DNA sequences in an unsupervised manner. We evaluate both quantitatively and qualitatively the learned latent representation for a supervised task of splice site classification. The quantitative evaluation is done under two different settings. Our experiments show that these representations can be used as features or priors in closely related tasks such as splice site classification. Further, in our qualitative analysis, we use a model attribution technique Integrated Gradients to infer significant sequence signatures influencing the classification accuracy. We show the identified splice signatures resemble well with the existing knowledge.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑