The rich biodiversity of India is home to an extensive range of indigenous plant and animal species, manyof which possess unique attributes with significant ecological, medicinal, and economic importance.Understanding the genetic makeup of these species is vital for their conservation and sustainableutilization. The focus of this paper is to present the Bharat Genome Database, a novel and comprehensivegenome database dedicated to the indigenous species of India. Our database provides a valuable resourcefor researchers, conservationists, and policymakers to study and preserve the genetic diversity of nativespecies.The Bharat Genome Database (BGD, bgdb.org) is a pioneering comparative genomics platform dedicatedto enhancing genomic research, training, and education related to species native to India. Featuring toolssuch as BLAST for sequence alignment and JBrowse and JBrowse2 for visualization of genomic data,BGD serves as a critical resource for scientists and educators alike. This paper introduces the database'sinfrastructure, bioinformatics tools, and potential to facilitate groundbreaking genomics and biodiversityconservation research.

Bacillus subtilis is an extensively characterized Gram-positive bacterium with broad applications in agriculture, biotechnology, and biocontrol. Strain SNUY1, known for its antifungal properties and use as an edible coating, has been completely sequenced and deposited in NCBI (accession: CP199918), making it an important model for genomic and functional characterization. Here, we present the MAYA pipeline, a fully automated, rule-based Snakemake workflow that integrates eighteen bioinformatic tools into a single end-to-end framework. MAYA accepts either raw Illumina paired-end FASTQ reads or pre-assembled FASTA sequences, utilizing per-rule Conda environment isolation to ensure strict version control and reproducibility across all constituent tools from quality control and assembly through functional annotation, plasmid typing, and detection of biosynthetic and antimicrobial resistance gene clusters. To benchmark pipeline accuracy across both input modes, MAYA was evaluated using Bacillus subtilis subsp. subtilis str. 168(AL009126 OZ025638). The FASTA assembly yielded a circularized 4,215,606 bp chromosome (N50: 4.21 Mb, 43.51% GC) with 98.28% CheckM completeness and 100.0% BUSCO ortholog recovery. Prokka annotation predicted 4214 CDSs (86 tRNAs, 30 rRNAs) in FASTA mode compared to 4313 CDSs (106 tRNAs, 25 rRNAs) in FASTQ mode, reflecting expected read fragmentation in unassembled data. DIAMOND mapping against SwissProt yielded 4135 hits (99.90% mean identity) in FASTA mode and 4170 hits (99.75% mean identity) in FASTQ mode. Both modes identified identical ST1 Multilocus Sequence Typing (MLST) profiles and six core AMR determinants (mphK, vmlR, rphC, aadK, satA_Bsub, and tet(L)), while antiSMASH detected 15 biosynthetic gene cluster (BGC) regions (22 protoclusters) in FASTA mode versus 16 regions (23 protoclusters) in FASTQ mode. MOB-suite resolved two non-mobilizable plasmids in the FASTQ assembly (bioproject: PRJNA76): pAC670 and a novel 1666 bp plasmid. Furthermore, nucleotide BLASTN analysis resolved a 76-CDS discrepancy between published NCBI annotations and MAYA/Prokka, demonstrating 95.5% (4081/4272) stringent sequence match and confirming that differences stem from gene-model predictions rather than omitted loci. To demonstrate the FASTA-mode application workflow, the complete SNUY1 (NZ_CP199918) genome assembly was processed through MAYA, yielding a single, contiguous chromosome of 4,196,931 base pairs with a GC content of 43.69% (98.28% CheckM, 100% BUSCO completeness). Annotation identified 4196 coding sequences, while antiSMASH profiling revealed 13 candidate biosynthetic gene cluster (BGC) regions comprising 20 protoclusters (including NRPS, lanthipeptides, terpenes, and sactipeptides), aligning with the strain’s biocontrol capabilities. AMR profiling identified six chromosomally encoded resistance genes (mphK, vmlR, rphC, aadK, satA_Bsub, and tet(L)). In silico mobility analysis via MOB-suite demonstrated that these sequences lack functional conjugation machinery and are classified as non-mobilizable. Executing in 3–4 h on an Intel Xeon server (128 GB RAM, 29 CPU cores), MAYA delivers an efficient, audit-ready framework for end-to-end bacterial genomic profiling.

Background and Objective: The dysbiosis of human gut microbiome has been increasingly seen to have a relation in the development of autoimmune diseases, with specific microbial signatures having causative association with specific conditions. Inflammatory bowel disease (IBD) is one such autoimmune ailment. This paper proposes a predictive tool that can identify the IBD status of an individual based on the composition of the gut microbiome using machine learning and AI agents driven techniques. The technology can strengthen the suspicion of a potential IBD diagnosis a patient may have based on their gut microbiome profile. Methods: The tool processes patient gut metagenome using integrated Kneaddata and MetaPhlAn to generate taxonomic profiles. These are fed into an XGBoost classifier to predict IBD or healthy status. Dysbiotic taxa are identified via Z-score and fold change. CrewAI delivers personalized probiotic recommendations based on diagnosis and dysbiosis. Results: The tuned XGBoost model achieved 86.6% accuracy. On validation using single ulcerative colitis sample, the tool correctly predicted IBD status but misclassified it as Crohn’s disease(possibly due to overlapping microbial signatures), identifying Faecalibacterium and Flavonifractor as dysbiotic taxa.The probiotic recommended was Faecalibacterium prausnitzii, backed with reasoning basedon scientific literature. Conclusions: Despite limited validation sample size, the high accuracy , correct IBD detection,dysbiosis analysis and elaborate probiotic recommendation suggest promising potential; further validation needed

Non-alcoholic fatty liver disease (NAFLD) is a globally prevalent hepatic condition caused by the buildup of fat in the liver. It is frequently associated with metabolic comorbidities such as hypertension, cardiovascular disease (CVD), and prediabetes. However, early detection remains challenging due to the asymptomatic progression, and existing primary diagnostic methods, such as imaging or liver biopsy, are often expensive and inaccessible in rural areas. This study proposes a two-stage, interpretable machine learning pipeline for the non-invasive and cost-effective prediction of NAFLD and its key comorbidities using routine clinical parameters. The NAFLD prediction model was developed using the XGBoost algorithm, trained on a hybrid dataset that combines real patient data with rule-based synthetic data generated by simulating clinically plausible cases. Upon NAFLD-positive prediction, three separate XGB models, trained on data labelled based on thresholds, assess individual risks for hypertension, cardiovascular disease, and prediabetes. Explainability is obtained using SHAP (SHapley Additive exPlanations), which provides insight into feature relevance, while biomarker radar plots help in the visual interpretation of comorbidities. A user-friendly Streamlit interface enables real-time interaction with the tool for potential clinical application. The NAFLD model demonstrated robust performance, while the models used for predicting comorbidities achieved perfect performance, which may be a reflection of the limited dataset size used in the second stage. This work underscores the potential of AI-driven tools in NAFLD diagnosis, particularly when combined with explainable AI methods.