2017/08/14 by Balázs Szalkai, Szalkai, Balazs, Vince Grolmusz +1
Biochemistry, Genetics and Molecular Biology · #Biomolecules (q-bio.BM) #FOS: Biological sciences #Genetics, Bioinformatics, and Biomedical Research #Genomics and Phylogenetic Studies #Machine Learning in Bioinformatics #Quantitative Methods (q-bio.QM)
paper · pdf · doi:10.48550/arxiv.1708.04103
openalex publication_date 2017/08/14 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Artificial intelligence (AI) tools are gaining more and more ground each year\nin bioinformatics. Learning algorithms can be taught easily by using the\nexisting enormous biological databases, and the resulting models can be used\nfor the high-quality classification of novel, un-categorized data in numerous\nareas, including biological sequence analysis. Here we introduce SECLAF, an\nartificial neural-net based biological sequence classifier framework, which\nuses the Tensorflow library of Google, Inc. By applying SECLAF for\nresidue-sequences, we have reported (Methods (2017),\nhttps://doi.org/10.1016/j.ymeth.2017.06.034) the most accurate multi-label\nprotein classifier to date (UniProt --into 698 classes-- AUC 99.99 %; Gene\nOntology --into 983 classes-- AUC 99.45 %). Our framework SECLAF can be applied\nfor other sequence classification tasks, as we describe in the present\ncontribution.\n Availability and implementation: The program SECLAF is implemented in Python,\nand is available for download, with example datasets at the website\nhttps://pitgroup.org/seclaf/. For Gene Ontology and UniProt based\nclassifications a webserver is also available at the address above.\n