Biology subjects

Kazmi, A.

Publications and source records attributed to Kazmi, A..

4 recordsLinked to original sources

bioRxiv2025

DEEPGEOSEARCH: LLM-POWERED SCHEMALESS RETRIEVALFOR BIOMEDICAL DATA DISCOVERY

Public biomedical repositories contain extensive, valuable datasets, yet identifying datasets that precisely match specific study requirements remains inefficient. Conventional keyword- and schema-based systems frequently fall short when queries encompass multiple biological and experimental facets. To address this, we developed DeepGEOSearch, an LLM-powered, schema-less retrieval system that interprets dataset text directly rather than relying solely on predefined metadata fields. The system extracts key study attributes, harmonizes terminology across sources, and ranks datasets by contextual alignment with natural-language queries while providing verifiable evidence for each match. In this work, we applied DeepGEOSearch for the Gene Expression Omnibus (GEO). The framework is repository-agnostic and can integrate new metadata sources without manual relabeling. This design enables complex compositional queries that existing tools cannot support. In evaluations against strong baselines using a curated evaluation benchmark comprising a mix of query complexities, DeepGEOSearch achieved more than 90% precision and best recall, with the largest performance gains observed on complex, real-world queries. DeepGEOSearch consistently identifies relevant datasets overlooked by conventional search tools, accelerating dataset discovery and improving reuse of public biomedical data.

bioinformatics↗

bioRxiv2025

A Hotspot Phosphorylation Site on SHP2 Drives Oncoprotein Activation and Drug Resistance

SHP2 is a phosphatase and a critical mediator of receptor tyrosine kinase (RTK)-driven RAS/mitogen-activated protein kinase (MAPK) signaling. Despite promising preclinical data, SHP2 inhibitors have shown minimal clinical efficacy, with no defined clinical mechanisms of primary resistance. Here, we elucidate phosphorylation of SHP2 at tyrosine 62 (pY62) as a hotspot phosphorylation site in the proteome and RTK-driven tumor types in patients. We demonstrate that SRC family kinases directly phosphorylate SHP2 at Y62, downstream of but not directly phosphorylated by RTKs. Using biochemical and biophysical analyses, we show that SHP2 Y62D enforces an open, active conformation, resulting in constitutive phosphatase activation that is sufficient to activate MAPK signaling and confer resistance to allosteric SHP2 inhibitors. These findings establish that SHP2 pY62 is a phosphorylation hotspot phenocopying mutational activation, a mechanism of primary resistance to SHP2 inhibitors, and a cancer drug target distinct from wildtype SHP2. Statement of significanceThis study identifies phosphorylation of SHP2 at tyrosine 62 (pY62) as a conserved mechanism of resistance to allosteric SHP2 inhibitors. By stabilizing an open, active SHP2 conformation, pY62 phenocopies oncogenic PTPN11 mutations and sustains MAPK signaling across cancer types. These findings redefine SHP2 inhibitor resistance as a phosphorylation-driven, target-intrinsic process, nominate pY62 as a potential biomarker for therapeutic response, and propose phosphorylated SHP2 as a distinct drug target.

cancer biology↗

bioRxiv2025

Beyond the Hype: The Complexity of Automated Cell Type Annotations with GPT-4

Recent research has shown the impressive capability of large language models like GPT-4 in various downstream tasks in single-cell data analysis. Among these tasks, cell type annotation remains particularly challenging, with researchers exploring various methods to improve accuracy and efficiency. While recent studies on GPT-like models have demonstrated annotation performance comparable to manual annotations, a significant gap remains in understanding their limitations and generalizability. In this work, we compare and evaluate the annotation performance of the GPT-4 model against traditional methods on nine randomly selected public single-cell RNA seq datasets from cellxgene, covering diverse tissue types. Our evaluation highlights the complexity of annotating cell types in single-cell data, revealing key differences between automated and manual approaches. We found specific cases where GPT-4 underperforms, demonstrating its limitations in certain contexts. We further introduce an automated approach to incorporate literature search using a RAG approach which enhances and outperforms GPT-4 cell type annotation when compared to traditional methods. We also introduce metrics based on taxonomic distance in the ontology tree to evaluate the granularity of the cell type annotations. To support future research, we also release an open-source Python package 1 that enables fully automated cell-type annotation of single-cell data using GPT-4 alongside other methods. The pipeline can take paper as an input and do cell type annotations on its own.

bioinformatics↗

bioRxiv2022

Bioinformatics techniques for efficient structure prediction of SARS-CoV-2 protein ORF7a via structure prediction approaches

Protein is the building block for all organisms. Protein structure prediction is always a complicated task in the field of proteomics. DNA and protein databases can find the primary sequence of the peptide chain and even similar sequences in different proteins. Mainly, there are two methodologies based on the presence or absence of a template for Protein structure prediction. Template-based structure prediction (threading and homology modeling) and Template-free structure prediction (ab initio). Numerous web-based servers that either use templates or do not can help us forecast the structure of proteins. In this current study, ORF7a, a transmembrane protein of the SARS-coronavirus, is predicted using Phyre2, IntFOLD, and Robetta. The protein sequence is straightforwardly entered into the sequence bar on all three web servers. Their findings provided information on the domain, the region with the disorder, the global and local quality score, the predicted structure, and the estimated error plot. Our study presents the structural details of the SARS-CoV protein ORF7a. This immunomodulatory component binds to immune cells and induces severe inflammatory reactions.

microbiology↗