Biology subjects

Dhir, R.

Publications and source records attributed to Dhir, R..

8 recordsLinked to original sources

bioRxiv2026

To Recover or Not to Recover: A Urinary Proteomic Perspective on Severe Acute Kidney Injury

Rationale & Objective Early prediction of renal non-recovery after acute kidney injury (AKI) could support risk-adapted follow-up and clinical decision-making. We evaluated whether a urinary fatty acid-binding protein 3 (FABP3)/apolipoprotein A-I (APOA1) ratio based on Total Protein Approach (TPA) quantification could discriminate renal recovery from non-recovery after severe AKI. Study Design Exploratory biomarker discovery study with outcome-blinded independent validation. Setting & Participants Critically ill adults with KDIGO stage 2 or 3 AKI. Urine was collected 48 hours after AKI diagnosis. The discovery cohort included 20 patients (10 recovered and 10 non-recovered), and the outcome-blinded validation cohort included 10 patients. New Predictor Urinary TPA-derived FABP3/APOA1 ratio (AKI^TPA index). Outcome Renal recovery at day 14, defined as survival with serum creatinine returning to within 0.3 mg/dL or 150% of baseline; patients not meeting these criteria were classified as having non-recovery. Analytical Approach Urinary proteins were analyzed by LC-MS/MS and quantified using TPA. Differential analysis identified FABP3 and APOA1 as oppositely associated with renal outcome. A predefined FABP3/APOA1 threshold of 1 classified patients as non-recovery-like (>1) or recovery-like (<1) and was subsequently applied to the outcome-blinded validation cohort. Results In the discovery cohort, the AKI^TPA index correctly classified 17 of 20 patients (accuracy, 85.0%; sensitivity, 80.0%; specificity, 90.0%; ROC AUC, 0.930). In the validation cohort, 9 patients were evaluable, of whom 7 were correctly classified (accuracy, 77.8%; sensitivity, 75.0%; specificity, 80.0%; ROC AUC, 0.950). Limitations The exploratory design and limited cohort size warrant further validation in larger, independent populations. Conclusions The urinary FABP3/APOA1 TPA ratio showed encouraging preliminary discrimination of renal recovery after severe AKI and warrants prospective clinical validation in larger, independent multicenter cohorts.

biochemistry↗

bioRxiv2026

Absolute quantitative proteomics guides patient-stratified drug repurposing in clear cell and papillary renal cell carcinoma.

BackgroundRenal cell carcinoma (RCC) is a highly heterogeneous disease in which distinct molecular subtypes exhibit characteristic genomic, metabolic, and microenvironmental features that influence therapeutic response. Substantial inter-patient variability exists within each subtype, resulting in markedly different clinical outcomes even among tumours of the same histological category. Proteomics provides a direct readout of tumour biology and pathway activity, complementing genomic information and enabling the identification of patient-specific actionable vulnerabilities. We applied a Total Protein Approach (TPA)-based prescriptomics framework that integrates absolute quantitative proteomics with curated drug-target knowledge to nominate patient-specific drug-repurposing options, positioned as coadjuvants to the prevailing standard of care. MethodsSeventeen human kidney tissue specimens, seven clear cell RCC (ccRCC), five papillary RCC (pRCC), and five normal adjacent tissues (NAT), were retrieved from the publicly available PRIDE repository (PXD023296) and reanalysed by TPA-based absolute quantification applied to previously acquired label-free LC-MS/MS data. Differential expression analysis between each tumour subtype and NAT identified subtype-specific upregulated proteins, wich were intersected with Therapeutic Target Database (TTD) to nominate FDA-approved drugs targeting dysregulated proteins as candidate repurposing strategies. ResultsccRCC and pRCC produced distinct proteome-wide upregulation profiles consistent with their known biological drivers. TPA index stratification nominated bempedoic acid (ACLY inhibitor) and tipiracil hydrochloride (TYMP inhibitor) as patient-stratified candidates for ccRCC, and auranofin (TXNRD1 inhibitor), bempedoic acid, and mipomersen (APOB-directed antisense oligonucleotide) for pRCC. ACLY was the only top-priority target shared across both subtypes, pointing to a candidate cross-subtype metabolic vulnerability. Secondary candidates emerged from protein-protein interaction network analysis in both subtypes.. ConclusionsThis study presents a quantitative proteomics framework for translating individual-patient proteomic dysregulation into coadjuvant drug-repurposing hypotheses across the principal RCC subtypes. By combining the TPA for absolute protein quantification with prescriptomics-guided drug-target mapping, we show that ccRCC and pRCC harbour distinct, individually stratifiable therapeutic vulnerabilities. These findings provide a proof-of-concept for proteomics-based treatment stratification in RCC and establish a scalable framework that, pending functional validation, could inform personalised therapeutic decision-making across RCC subtypes.

biochemistry↗

bioRxiv2025

Self-supervised learning enables robust microbiome predictions in data-limited and cross-cohort settings

The gut microbiome plays a crucial role in human health, but machine learning applications in this field face significant challenges, including limited labeled data availability, high dimensionality, and batch effects across different cohorts. To address these limitations, we developed representation learning models for gut microbiome metagenomic data, drawing inspiration from foundation models approaches based on self-supervised and transfer learning principles. By leveraging a large collection of 85,364 metagenomic samples, we implemented multiple self-supervised learning methods, including masked autoencoders with varying masking rates and adapted single-cell RNA sequencing models (scVI and scGPT), to generate embeddings from bacterial abundance profiles. These learned representations demonstrated significant advantages over raw bacterial abundances in two key scenarios: first, when training predictive models with very limited labeled data, improving prediction performance for age (r = 0.14 vs. 0.06), BMI (r = 0.16 vs. 0.11), visceral fat mass (r = 0.25 vs. 0.18), and drug usage classification (PR-AUC = 0.81 vs. 0.73); and second, when generalizing predictions across different cohorts, consistently outperforming models based on raw abundances in cross-dataset evaluation. Our approach provides a valuable framework for leveraging self-supervised representation learning to overcome the data limitations inherent in microbiome research, potentially enabling more robust and generalizable machine learning applications in this field.

systems biology↗

bioRxiv2025

Comprehensive Survey of Gut Microbiome Associations with Health Conditions in the Human Phenotype Project

The human gut microbiome is increasingly implicated with diverse health conditions, highlighting the importance of developing comprehensive resources to systematically map these associations. In this study, we leveraged the Human Phenotype Project (HPP) 10K cohort, a large, deeply phenotyped population study cohort with extensive metagenomic profiling, to explore associations between the gut microbiome and 37 health indications. These include conditions with established or emerging links to the microbiome, such as obesity, diabetes, hyperlipidemia, metabolic syndrome, inflammation, liver disease, cardiovascular and kidney conditions, immune and allergic diseases, gastrointestinal symptoms, mental health, and sleep-related traits. Using a curated set of approximately 200 refined phenotypic features and analyzing 1,184 microbial species, we performed robust statistical association analyses. We observed significant enrichment of associations in 14 health indications, with most associations reflecting increased microbial abundance in favorable health states, thereby suggesting potential microbial targets for intervention. This work introduces a publicly available, high-resolution resource to facilitate future research and support the development of microbiome-informed health strategies.

genomics↗

bioRxiv2023

A study of machine learning techniques for Automated Karyotyping System

Genetic abnormalities constitute a considerable share of all the existing societal healthcare issues. There has been a dire need for the automation of chromosomal analysis, hence supporting laboratory workers in effective classification and identifying such abnormalities. Nevertheless, with many modern image processing techniques, like Karyotyping, improved the life expectancy and the quality of life of such cases. The standard image-based analysis procedures include Pre-processing, Segmentation, Feature extraction, and Classification of images. When explicitly considering Karyotyping, the processes of Segmentation and Classification of chromosomes have been the most complex, with much existing literature focusing on the same. Various model-based machine learning models have proven to be highly effective in solving existing issues and building an artificial intelligence-based, autonomous-centric karyotyping system. An autonomous Karyotyping System will connect the pre-processing, Segmentation, and classification of metaphase images. The review focuses on machine learning-based algorithms for efficient classification accuracy. The study has the sole motive of moving towards an effective classification method for karyotype metaphase images, which will eventually predict the fetuss abnormalities more effectively. The studys results shall benefit future researchers working in this area.

bioengineering↗

bioRxiv2023

Cyanobacteria newly isolated from marine volcanic seeps display rapid sinking and robust, high density growth

Cyanobacteria are photosynthetic organisms that play important roles in carbon cycling as well as promising bioproduction chassis. Here, we isolate two novel cyanobacteria, UTEX 3221 and UTEX 3222, from a unique marine environment with naturally elevated CO2. We describe complete genome sequences for both isolates and, focusing on UTEX 3222 due to its planktonic growth in liquid, characterize biotechnologically-relevant growth and biomass characteristics. UTEX 3222 outpaces other fast-growing model strains on solid medium. It can double every 2.35 hours in a liquid medium and grows to high density (>31g/L biomass dry weight) in batch culture, nearly double that of Synechococcus sp. PCC 11901, whose high-density growth was recently reported. In addition, UTEX 3222 sinks readily, settling more quickly than other fast-growing strains, suggesting improved de-watering of UTEX 3222 biomass. This settling behavior can be explained in part by larger cell volume. These traits may make UTEX 3222 a compelling choice for photosynthetic bioproduction from CO2. Overall, we find that bio-prospecting in environments with naturally elevated CO2 may uncover novel CO2-metabolizing organisms with unique characteristics.

microbiology↗

bioRxiv2021

Pathway centered analysis to guide clinical decision-making in precision medicine

Changes in the human proteome caused by disease before, during and after medical care is phenotype-dependent, so the proteome of each individual at any time point is a snapshot of the bodys response to disease and to disease treatment. Here, we introduce a new concept named differential Personal Pathway index (dPPi). This tool extracts and summates comprehensive disease-specific information contained within an individuals proteome as a holistic way to follow the response to disease and medical care over time. We demonstrate the principle of the dPPi algorithm on proteins found in urine from patients suffering from neoplasia of the bladder. The relevance of the dPPi results to the individual clinical cases is described. The dPPi concept can be extended to other malignant and non-malignant diseases, and to other types of biopsies, such as plasma, serum or saliva. We envision the dPPi as a tool for clinical decision-making in precision medicine.

biochemistry↗

bioRxiv2021

6mer Seed Toxicity and Platinum Resistance in Ovarian Cancer

Numerous micro(mi)RNAs (short noncoding RNAs that negatively regulate gene expression) have been linked to platinum (Pt) sensitivity and resistance in ovarian cancer (OC). miRNA activity occurs when the guide strand of the miRNA, with its seed sequence (pos. 2-7/8), is loaded into the RNA induced silencing complex (RISC) and targets complementary short seed matches in the 3 untranslated region of mRNAs. Toxic seeds, targeting genes critical for cancer cell survival, have been found in tumor suppressive miRNAs. Many si- and shRNAs can also kill cancer cells via toxic seeds, the most toxic carrying G-rich 6mer seed sequences. We now show that treatment of OC cells with Pt leads to an increase in RISC-bound miRNAs carrying toxic 6mer seeds and a decrease in miRNAs with nontoxic seeds. Pt-resistant cells did not exhibit this toxicity shift but retained sensitivity to cell death mediated by siRNAs carrying toxic 6mer seeds. Analysis of RISC-bound miRNAs in OC patients revealed that the ratio between miRNAs with toxic versus miRNAs with nontoxic seeds was predictive of treatment outcome. Application of the 6mer seed toxicity concept to cancer relevant miRNAs provides a new framework for understanding and predicting cancer therapy responses.

cancer biology↗