24 March 2018 in Bioinformatics
Bacterial resistance to antibiotics is a growing concern. Antimicrobial peptides (AMPs), natural components of innate immunity, are popular targets for developing new drugs. Machine learning methods are now commonly adopted by wet-laboratory researchers to screen for promising candidates. In this work, we utilize deep learning to recognize antimicrobial activity. We propose a neural network model with convolutional and recurrent layers that leverage primary sequence composition. Results show that the proposed model outperforms state-of-the-art classification models on a comprehensive dataset. By utilizing the embedding weights, we also present a reduced-alphabet representation and show that reasonable AMP recognition can be maintained using nine amino acid types. Models and datasets are made freely available through the Antimicrobial Peptide Scanner vr.2 web server at www.ampscanner.com. Supplementary data are available at Bioinformatics online.
3 May 2016 in Nucleic Acids Research
Defining syntenic relationships among orthologous gene clusters is a frequent undertaking of biologists studying organismal evolution through comparative genomic approaches. With the increasing availability of genome data made possible through next-generation sequencing technology, there is a growing need for user-friendly tools capable of assessing synteny. Here we present SimpleSynteny, a new web-based platform capable of directly interrogating collinearity of local genomic neighbors across multiple species in a targeted manner. SimpleSynteny provides a pipeline for evaluating the synteny of a preselected set of gene targets across multiple organismal genomes. An emphasis has been placed on ease-of-use, and users are only required to submit FASTA files for their genomes and genes of interest. SimpleSynteny then guides the user through an iterative process of exploring and customizing genomes individually before combining them into a final high-resolution figure. Because the process is iterative, it allows the user to customize the organization of multiple contigs and incorporate knowledge from additional sources, rather than forcing complete dependence on the computational predictions. Additional tools are provided to help the user identify which contigs in a genome assembly contain gene targets and to optimize analyses of circular genomes. SimpleSynteny is freely available at: http://www.SimpleSynteny.com.
1 April 2019 in Journal of Allergy and Clinical Immunology
Caspase activation and recruitment domain 11 (CARD11) encodes a scaffold protein in lymphocytes that links antigen receptor engagement with downstream signaling to nuclear factor κB, c-Jun N-terminal kinase, and mechanistic target of rapamycin complex 1. Germline CARD11 mutations cause several distinct primary immune disorders in human subjects, including severe combined immune deficiency (biallelic null mutations), B-cell expansion with nuclear factor κB and T-cell anergy (heterozygous, gain-of-function mutations), and severe atopic disease (loss-of-function, heterozygous, dominant interfering mutations), which has focused attention on CARD11 mutations discovered by using whole-exome sequencing. We sought to determine the molecular actions of an extended allelic series of CARD11 and to characterize the expanding range of clinical phenotypes associated with heterozygous CARD11 loss-of-function alleles. Cell transfections and primary T-cell assays were used to evaluate signaling and function of CARD11 variants. Here we report on an expanded cohort of patients harboring novel heterozygous CARD11 mutations that extend beyond atopy to include other immunologic phenotypes not previously associated with CARD11 mutations. In addition to (and sometimes excluding) severe atopy, heterozygous missense and indel mutations in CARD11 presented with immunologic phenotypes similar to those observed in signal transducer and activator of transcription 3 loss of function, dedicator of cytokinesis 8 deficiency, common variable immunodeficiency, neutropenia, and immune dysregulation, polyendocrinopathy, enteropathy, X-linked-like syndrome. Pathogenic variants exhibited dominant negative activity and were largely confined to the CARD or coiled-coil domains of the CARD11 protein. These results illuminate a broader phenotypic spectrum associated with CARD11 mutations in human subjects and underscore the need for functional studies to demonstrate that rare gene variants encountered in expected and unexpected phenotypes must nonetheless be validated for pathogenic activity.
20 January 2021 in Bioinformatics
Automatic phenotype concept recognition from unstructured text remains a challenging task in biomedical text mining research. Previous works that address the task typically use dictionary-based matching methods, which can achieve high precision but suffer from lower recall. Recently, machine learning-based methods have been proposed to identify biomedical concepts, which can recognize more unseen concept synonyms by automatic feature learning. However, most methods require large corpora of manually annotated data for model training, which is difficult to obtain due to the high cost of human annotation. In this article, we propose PhenoTagger, a hybrid method that combines both dictionary and machine learning-based methods to recognize Human Phenotype Ontology (HPO) concepts in unstructured biomedical text. We first use all concepts and synonyms in HPO to construct a dictionary, which is then used to automatically build a distantly supervised training dataset for machine learning. Next, a cutting-edge deep learning model is trained to classify each candidate phrase (n-gram from input sentence) into a corresponding concept label. Finally, the dictionary and machine learning-based prediction results are combined for improved performance. Our method is validated with two HPO corpora, and the results show that PhenoTagger compares favorably to previous methods. In addition, to demonstrate the generalizability of our method, we retrained PhenoTagger using the disease ontology MEDIC for disease concept recognition to investigate the effect of training on different ontologies. Experimental results on the NCBI disease corpus show that PhenoTagger without requiring manually annotated training data achieves competitive performance as compared with state-of-the-art supervised methods. The source code, API information and data for PhenoTagger are freely available at https://github.com/ncbi-nlp/PhenoTagger. Supplementary data are available at Bioinformatics online.
8 October 2013 in BioEssays