Institute of Computer Science, Martin Luther University Halle-Wittenberg, Halle (Saale), Germany.
Institute for Biosafety in Plant Biotechnology, Julius Kühn-Institut (JKI), Quedlinburg, Germany.
Bioinformatics. 2019 Nov 1;35(22):4812-4814. doi: 10.1093/bioinformatics/btz507.
Statistical dependencies are present in a variety of sequence data, but are not discernible from traditional sequence logos. Here, we present the R package DepLogo for visualizing inter-position dependencies in aligned sequence data as dependency logos. Dependency logos make dependency structures, which correspond to regular co-occurrences of symbols at dependent positions, visually perceptible. To this end, sequences are partitioned based on their symbols at highly dependent positions as measured by mutual information, and each partition obtains its own visual representation. We illustrate the utility of the DepLogo package in several use cases generating dependency logos from DNA, RNA and protein sequences.
The DepLogo R package is available from CRAN and its source code is available at https://github.com/Jstacs/DepLogo.
Supplementary data are available at Bioinformatics online.
各种序列数据中都存在统计相关性,但传统的序列 logo 无法识别这些相关性。在这里,我们介绍了 R 包 DepLogo,用于将对齐序列数据中的位置间相关性可视化为依赖 logo。依赖 logo 可使依赖结构(对应于符号在依赖位置上的规则共现)可视化。为此,序列根据其在高度依赖位置上的符号进行分区,这些符号是通过互信息来衡量的,每个分区都获得自己的可视化表示。我们通过从 DNA、RNA 和蛋白质序列生成依赖 logo 的几个用例说明了 DepLogo 包的实用性。
DepLogo R 包可从 CRAN 获得,其源代码可在 https://github.com/Jstacs/DepLogo 上获得。
补充数据可在生物信息学在线获得。