BackgroundThe availability of sequence data of human pathogenic fungi generates opportunities to develop Bioinformatics tools and resources for vaccine development towards benefitting at-risk patients.DescriptionWe have developed a fungal adhesin predictor and an immunoinformatics database with predicted adhesins. Based on literature search and domain analysis, we prepared a positive dataset comprising adhesin protein sequences from human fungal pathogens Candida albicans, Candida glabrata, Aspergillus fumigatus, Coccidioides immitis, Coccidioides posadasii, Histoplasma capsulatum, Blastomyces dermatitidis, Pneumocystis carinii, Pneumocystis jirovecii and Paracoccidioides brasiliensis. The negative dataset consisted of proteins with high probability to function intracellularly. We have used 3945 compositional properties including frequencies of mono, doublet, triplet, and multiplets of amino acids and hydrophobic properties as input features of protein sequences to Support Vector Machine. Best classifiers were identified through an exhaustive search of 588 parameters and meeting the criteria of best Mathews Correlation Coefficient and lowest coefficient of variation among the 3 fold cross validation datasets. The "FungalRV adhesin predictor" was built on three models whose average Mathews Correlation Coefficient was in the range 0.89-0.90 and its coefficient of variation across three fold cross validation datasets in the range 1.2% - 2.74% at threshold score of 0. We obtained an overall MCC value of 0.8702 considering all 8 pathogens, namely, C. albicans, C. glabrata, A. fumigatus, B. dermatitidis, C. immitis, C. posadasii, H. capsulatum and P. brasiliensis thus showing high sensitivity and specificity at a threshold of 0.511. In case of P. brasiliensis the algorithm achieved a sensitivity of 66.67%. A total of 307 fungal adhesins and adhesin like proteins were predicted from the entire proteomes of eight human pathogenic fungal species. The immunoinformatics analysis data on these proteins were organized for easy user interface analysis. A Web interface was developed for analysis by users. The predicted adhesin sequences were processed through 18 immunoinformatics algorithms and these data have been organized into MySQL backend. A user friendly interface has been developed for experimental researchers for retrieving information from the database.ConclusionFungalRV webserver facilitating the discovery process for novel human pathogenic fungal adhesin vaccine has been developed.

No abstract

We performed integrative analysis of genes associated with type 2 Diabetes Mellitus (T2DM) associated complications by automated text mining with manual curation and also gene expression analysis from Gene Expression Omnibus. They were analysed for pathogenic or protective role, trends, interaction with risk factors, Gene Ontology enrichment and tissue wise differential expression. The database T2DiACoD houses 650 genes, and 34 microRNAs associated with T2DM complications. Seven genes AGER, TNFRSF11B, CRK, PON1, ADIPOQ, CRP and NOS3 are associated with all 5 complications. Several genes are studied in multiple years in all complications with high proportion in cardiovascular (75.8%) and atherosclerosis (51.3%). T2DM Patients’ skeletal muscle tissues showed high fold change in differentially expressed genes. Among the differentially expressed genes, VEGFA is associated with several complications of T2DM. A few genes ACE2, ADCYAP1, HDAC4, NCF1, NFE2L2, OSM, SMAD1, TGFB1, BDNF, SYVN1, TXNIP, CD36, CYP2J2, NLRP3 with details of protective role are catalogued. Obesity is clearly a dominant risk factor interacting with the genes of T2DM complications followed by inflammation, diet and stress to variable extents. This information emerging from the integrative approach used in this work could benefit further therapeutic approaches. The T2DiACoD is available at www.http://t2diacod.igib.res.in/.

SARS-CoV2 pandemic exposed the limitations of artificial intelligence based medical imaging systems. Earlier in the pandemic, the absence of sufficient training data prevented effective deep learning (DL) solutions for the diagnosis of COVID-19 based on X-Ray data. Here, addressing the lacunae in existing literature and algorithms with the paucity of initial training data; we describe CovBaseAI, an explainable tool using an ensemble of three DL models and an expert decision system (EDS) for COVID-Pneumonia diagnosis, trained entirely on pre-COVID-19 datasets. The performance and explainability of CovBaseAI was primarily validated on two independent datasets. Firstly, 1401 randomly selected CxR from an Indian quarantine center to assess effectiveness in excluding radiological COVID-Pneumonia requiring higher care. Second, curated dataset; 434 RT-PCR positive cases and 471 non-COVID/Normal historical scans, to assess performance in advanced medical settings. CovBaseAI had an accuracy of 87% with a negative predictive value of 98% in the quarantine-center data. However, sensitivity was 0.66–0.90 taking RT-PCR/radiologist opinion as ground truth. This work provides new insights on the usage of EDS with DL methods and the ability of algorithms to confidently predict COVID-Pneumonia while reinforcing the established learning; that benchmarking based on RT-PCR may not serve as reliable ground truth in radiological diagnosis. Such tools can pave the path for multi-modal high throughput detection of COVID-Pneumonia in screening and referral.