A Comparison of Function Annotation Methods over Multiple Meta-omics Data of Microbiota
Cristopher Reyes Loaiciga
School Life Sciences and Biotechnology, Shanghai Jiaotong University, 200240
Genomics and Applied Biology, 2020, Vol. 11, No. 1
Received: 02 Mar., 2020 Accepted: 30 Mar., 2020 Published: 15 Apr., 2020
© 2020 BioPublisher Publishing Platform
This is an open access article published under the terms of the Creative Commons Attribution License
, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
The functional annotation of meta-omics DNA or Protein sequences usually only gives function labels to a fraction of any given dataset, which has created huge gaps of newly discovered proteins, that remain, uncharacterized. Some methods are highly popular because they can handle big sets of data, however, the results may vary from one to another, leaving the question of which one to use. Three commonly used software (OMA, EggNOG and InterproScan) were selected to evaluate and assess their quality of prediction over ten species of well annotated genomes of bacteria, 243 299 genes and 3 418 protein groups predicted from human gut metagenomics and meta-proteomics samples. Here both the coverage and the quality of the annotation were assessed, adding two additional baseline methods: one based on BLAST and one based on the frequency of each function. We found that InterproScan provided better results in terms of overall coverage, but no method outperformed the others in all the ontology categories, as EggNOG constantly performed higher than in InterProScan for cellular component ontologies. Researchers should be aware of which tool is better suited for their own particular case and interests.
Software evaluation; Gene ontology; Function annotation; Human gut microbiome
Genomics and Applied Biology
• Volume 11