Biology subjects

Hossain, M. N.

Publications and source records attributed to Hossain, M. N..

4 recordsLinked to original sources

bioRxiv2025

Nonsense mutations can also increase mRNA levels

Nonsense mutations are often associated with reduced mRNA levels, as premature translation termination can lead to the activation of nonsense-mediated mRNA decay (NMD). To examine how positional context influences these outcomes, we introduced premature translation termination codons (PTCs) at 15 locations within the coding region of a GFP reporter gene in Schizosaccharomyces pombe. PTCs in the first third of the coding region (codons 1-88) consistently led to reduced mRNA levels. In contrast, most downstream PTCs (codons 108-231) showed modest or minimal reductions, and several were associated with increased mRNA levels relative to the PTC-less control transcript. Measurement of transcript stability for one such variant indicated that the increased abundance was not attributable to decreased turnover. Deletion of UPF1 in wild-type cells elevated the levels of transcripts that were reduced and, unexpectedly, further increased the abundance of transcripts that exceeded the control level. In spliced versions of these constructs, downstream PTCs generally reduced mRNA levels regardless of exon junction position; additionally, for some early PTCs, splicing appeared to suppress rather than enhance mRNA reduction. Overall, these observations indicate that an unexpected consequence of nonsense mutations can be increased mRNA levels. These findings may aid in the interpretation of the effects of nonsense mutations on mRNA abundance beyond the predictions of current NMD models and may also help in the design of eukaryotic gene-expression constructs.

molecular biology↗

bioRxiv2025

BioFuse: An Embedding Fusion Framework for Biomedical Foundation Models

The biomedical field has witnessed a surge in pre-trained foundation mod-els excelling in specific sub-domains such as radiology and histopathology. While integrating these models promises a more comprehensive understand-ing of biomedical data, it poses challenges in model compatibility and feature fusion. We present BioFuse, a novel open-source framework designed to generate optimised biomedical embeddings. BioFuse utilises a pool of 9 state-of-the-art foundation models to create task-specific embeddings. It employs grid search to automatically identify the optimal combination of models, fusing their embeddings through vector concatenation. On the MedMNIST+ benchmark, using XGBoost as the downstream classifier, Bio-Fuse outperforms several existing methods, achieving SOTA AUC in 5/12 datasets while maintaining near-SOTA performance across most remain-ing datasets. Remarkably, our experiments reveal unexpected cross-modal capabilities, with histopathology and radiology models showing strong performance when applied to other imaging modalities. BioFuse features a high-level API1 for immediate deployment and an extensible architecture to incorporate future models and fusion techniques. We anticipate BioFuse will not only enhance the utility of foundation models in biomedicine but also open new avenues for uncovering cross-modal relationships in biomedical data.

bioinformatics↗

bioRxiv2019

Conglomeration of highly antigenic nucleoproteins to inaugurate a heterosubtypic next generation vaccine candidate against Arenaviridae family

AbstractArenaviral infections often resulting in lethal Hemorrhagic Fevers (HF) affect primarily African and South American regions. To date, there is no FDA-approved licensed vaccine against arenaviruses and treatments have been limited to supportive therapies. Hence, the study was employed to design a highly immunogenic heterosubtypic vaccine candidate against Arenaviridae family. The whole proteomes of Lassa virus (LASV), Lymphocytic Chorio Meningitis Virus (LCMV), Lujo virus and Guanarito virus were retrieved from NCBI database and assessed to determine the most antigenic viral proteins. Only the conserved sequences were used for T cell and B cell epitope prediction to ensure protective response against a wide range of viral strains. For each virus, nucleoproteins were identified as most antigenic which generated a plethora of antigenic epitopes. The proposed epitopes were highly conserved (up to 100%) and showed high cumulative population coverage. Moreover, results revealed that among the top epitopes, T cell epitope GWPYIGSRS were conserved in Argentine mammarenavirus (Junin virus) and Brazilian mammarenavirus (Sabia virus), while B cell epitope NLLYKICLSG were conserved in Bolivian mammarenavirus (Machupo virus) and Brazilian mammarenavirus (Sabia virus), indicating the possibility of final vaccine constructs to confer broad range immunity in the host. A total 3 constructs were designed by the combination of top epitopes from each protein along with suitable adjuvant and linkers. Different physicochemical properties revealed the superiority of construct V1 in terms of safety and efficacy. Docking analysis of the refined vaccine structure with different MHC molecules and human immune receptors were also biologically significant. The vaccine receptor complex (V1-TLR3) showed minimal deformability at molecular level. Moreover, construct V1 was compatible for insertion into pET28a(+) vector and heterologous cloning in E. coli srain K12. However, the results were based on different sequence analysis and various immune databases. Further wet lab based studies using model animals are highly recommended for the experimental validation of the designed vaccine candidates.

bioinformatics↗

bioRxiv2019

Contriving a chimeric polyvalent vaccine to prevent infections caused by Herpes Simplex Virus (Type-1 and Type-2): an exploratory immunoinformatic approach

Herpes simplex virus type 1 (HSV-1) and 2 (HSV-2) cause a variety of infections including oral-facial infections, genital herpes, herpes keratitis, cutaneous infection and so on. To date, FDA-approved licensed HSV vaccine is not available yet. Hence, the study was conducted to identify and characterize an effective epitope based polyvalent vaccine against both types of Herpes Simplex Virus through targeting six viral proteins. The selected proteins were retrieved from viralzone and assessed to design highly antigenic epitopes by binding analyses of the peptides with MHC class-I and class-II molecules, antigenicity screening, transmembrane topology screening, allergenicity and toxicity assessment, population coverage analysis and molecular docking approach. The final vaccine was constructed by the combination of top CTL, HTL and BCL epitopes from each protein along with suitable adjuvant and linkers. Physicochemical and secondary structure analysis, disulfide engineering, molecular dynamic simulation and codon adaptation were further employed to develop a unique multi-epitope peptide vaccine. Docking analysis of the refined vaccine structure with different MHC molecules and human immune TLR-2 receptor demonstrated higher interaction. Complexed structure of the modeled vaccine and TLR-2 showed minimal deformability at molecular level. Moreover, translational potency and microbial expression of the modeled vaccine was analyzed with pET28a(+) vector for E. coli strain strain K12. The study enabled design of a novel chimeric polyvalent vaccine to confer broad range immunity against both HSV serotypes. However, further wet lab based research using model animals are highly recommended to experimentally validate our findings.

bioinformatics↗