Sreeram Chandra Murthy Peela
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Sreeram Chandra Murthy Peela has a total of 79 co-authors. Here are the top co-authors by number of shared publications:
- Sujatha Sistla
Department of Microbiology, Jawaharlal Institute of Postgraduate Medical Education and Research, Puducherry, India
7 shared publications - Gaurav Ahuja
Department of Computational Biology, Indraprastha Institute of Information Technology-Delhi (IIIT-Delhi), Okhla, Phase III, New Delhi, 110020, India
3 shared publications - Nilabja Bhattacharjee
Department of Computational Biology, Indraprastha Institute of Information Technology - Delhi (IIITD), New Delhi, India
3 shared publications - Abhishek Halder
Department of Computer Sciences and Engineering, Indraprastha Institute of Information Technology - Delhi (IIITD), New Delhi, India
2 shared publications - Bernadette Mathew
Department of Computational Biology, Indraprastha Institute of Information Technology - Delhi (IIITD), New Delhi, India
2 shared publications - Debarka Sengupta
Department of Computational Biology, Indraprastha Institute of Information Technology-Delhi (IIIT-Delhi), Okhla, Phase III, New Delhi, 110020, India
2 shared publications
Latest Publications
- Open Access
- 8 June 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
Gene dependency-informed inference of response to targeted cancer therapies
Targeted therapies such as small-molecule inhibitors act by blocking proteins essential for cancer cell survival, yet omics-based modeling of drug sensitivity often lacks mechanistic grounding. We present FORGE (Factorization Of Response and Gene Essentiality), a joint matrix factorization framework that co-models drug response and target gene dependency to enable biologically informed stratification of treatment groups. FORGE derives a Benefit Score from basal gene expression to estimate therapeutic potential. In unseen cell lines treated with erlotinib, FORGE achieves high concordance for dependency (0.69) and IC50 (0.62), with benefit score stratification showing increased dependency and decreased IC50 across quartiles. Joint modeling improves predictive performance over single-task approaches and enhances agreement between gene-level effects (p = 0.039). Validation across independent datasets shows that higher benefit scores associate with tumor regression in patient-derived xenografts and with predicted drug susceptibility in the Tahoe-100M dataset. Mechanistic analyses further identify gene programs underlying drug susceptibility.
- 8 June 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
Open AccessGene dependency-informed inference of response to targeted cancer therapies
Targeted therapies such as small-molecule inhibitors act by blocking proteins essential for cancer cell survival, yet omics-based modeling of drug sensitivity often lacks mechanistic grounding. We present FORGE (Factorization Of Response and Gene Essentiality), a joint matrix factorization framework that co-models drug response and target gene dependency to enable biologically informed stratification of treatment groups. FORGE derives a Benefit Score from basal gene expression to estimate therapeutic potential. In unseen cell lines treated with erlotinib, FORGE achieves high concordance for dependency (0.69) and IC50 (0.62), with benefit score stratification showing increased dependency and decreased IC50 across quartiles. Joint modeling improves predictive performance over single-task approaches and enhances agreement between gene-level effects (p = 0.039). Validation across independent datasets shows that higher benefit scores associate with tumor regression in patient-derived xenografts and with predicted drug susceptibility in the Tahoe-100M dataset. Mechanistic analyses further identify gene programs underlying drug susceptibility.
- Open Access
A knowledge graph-guided framework for multitask single-cell analysis
Connecting cell identity to biological function is central to understanding how tumors change during treatment. Here, we introduce scKNIFE, a knowledge-guided nonnegative matrix factorization framework that integrates gene expression with a heterogeneous graph of functional annotations, metabolic reactions and metabolites, and signaling interactions. scKNIFE jointly learns cellular representations and biological-entity loadings, linking the factors that organize cells to their constituent genes and connected functions. Across breast-cancer cohorts, the learned representations retained broad cellular organization without using per-cell identity labels as factorization targets. scKNIFE identified baseline tumor programs associated with subsequent clinical response or T-cell clonotype expansion and resolved treatment-associated functional variation within immune and stromal populations. In a seven-patient triple-negative breast cancer atlas, analysis alongside tissue metabolomics identified distinct trajectories of inferred fatty acid activation and carnitine-associated activities, including higher myeloid L-octanoylcarnitine scores at biopsy in all seven patients. Communication analysis further resolved an endothelial-perivascular PDGF connection with lower inferred scores at surgery in six patients. scKNIFE brings cellular identity, metabolic programs and candidate intercellular communication into one representation, enabling functional interpretation of tumor heterogeneity across patients and treatment states.
Open AccessA knowledge graph-guided framework for multitask single-cell analysis
Connecting cell identity to biological function is central to understanding how tumors change during treatment. Here, we introduce scKNIFE, a knowledge-guided nonnegative matrix factorization framework that integrates gene expression with a heterogeneous graph of functional annotations, metabolic reactions and metabolites, and signaling interactions. scKNIFE jointly learns cellular representations and biological-entity loadings, linking the factors that organize cells to their constituent genes and connected functions. Across breast-cancer cohorts, the learned representations retained broad cellular organization without using per-cell identity labels as factorization targets. scKNIFE identified baseline tumor programs associated with subsequent clinical response or T-cell clonotype expansion and resolved treatment-associated functional variation within immune and stromal populations. In a seven-patient triple-negative breast cancer atlas, analysis alongside tissue metabolomics identified distinct trajectories of inferred fatty acid activation and carnitine-associated activities, including higher myeloid L-octanoylcarnitine scores at biopsy in all seven patients. Communication analysis further resolved an endothelial-perivascular PDGF connection with lower inferred scores at surgery in six patients. scKNIFE brings cellular identity, metabolic programs and candidate intercellular communication into one representation, enabling functional interpretation of tumor heterogeneity across patients and treatment states.
- Open Access
AI for Computational Biology: Highlights from the first BioAI Hackathon at University of Warsaw
The BioAI Hackathon at the Centre of New Technologies at the University of Warsaw convened 43 international researchers to collaboratively explore artificial intelligence (AI) approaches for solving complex challenges in computational biology. Nine interdisciplinary and multi-institutional teams addressed the following problems: disease-gene prioritization, microbiome analysis, drug-protein interactions, alternative splicing prediction, chromatin architecture study and toxicological profiling. Using cutting-edge tools such as graph neural networks (GNNs), large language models (LLMs), and multi-omics integration frameworks, participants developed scalable and reproducible analytical pipelines. The results include: a disease gene prioritization framework using GNNs, a microbiome dynamics analysis for poultry health prediction and the construction of chromatin structure-aware regulatory networks. All projects follow the open science principles and display translational potential. This hackathon underscores the transformative role of AI in biomedicine and the value of collaborative, time-bounded innovation for accelerating discovery in life sciences. All projects are publicly available on GitHub: https://github.com/SFGLab
Open AccessAI for Computational Biology: Highlights from the first BioAI Hackathon at University of Warsaw
The BioAI Hackathon at the Centre of New Technologies at the University of Warsaw convened 43 international researchers to collaboratively explore artificial intelligence (AI) approaches for solving complex challenges in computational biology. Nine interdisciplinary and multi-institutional teams addressed the following problems: disease-gene prioritization, microbiome analysis, drug-protein interactions, alternative splicing prediction, chromatin architecture study and toxicological profiling. Using cutting-edge tools such as graph neural networks (GNNs), large language models (LLMs), and multi-omics integration frameworks, participants developed scalable and reproducible analytical pipelines. The results include: a disease gene prioritization framework using GNNs, a microbiome dynamics analysis for poultry health prediction and the construction of chromatin structure-aware regulatory networks. All projects follow the open science principles and display translational potential. This hackathon underscores the transformative role of AI in biomedicine and the value of collaborative, time-bounded innovation for accelerating discovery in life sciences. All projects are publicly available on GitHub: https://github.com/SFGLab
- Open Access
Mechanism-aware inference of response to targeted cancer therapies
Targeted therapies like small-molecule inhibitors often work by blocking proteins that cancer cells rely on for survival. Omics based modeling of drug sensitivity alone lack mechanistic grounding. We propose FORGE (Factorization Of Response and Gene Essentiality) a simple yet powerful joint matrix factorization framework that co-models drug response and target gene essentiality, enabling the stratification of promising treatment groups for targeted therapy consideration. FORGE also provides Benefit Score-a predictive score that estimates treatment efficacy from basal gene expression profiles. We validated the predictive performance of FORGE across multiple targeted therapies, including Erlotinib (EGFR inhibitor) and Daporinad (NAMPT inhibitor). Our meta-analysis of large scale in-vitro studies underscores FORGE's ability to identify common determinants of drug vulnerabilities and target gene essentiality. Such convergences were not observed when treatment vulnerabilities and gene essentialities were modeled independently. We also demonstrated the universality of Erlotinib Benefit Scores by transferring transformations learned from high-throughput drug response studies across other published datasets, including the TAHOE-100M single-cell perturbation atlas and patient-derived xenograft studies. FORGE successfully identified key regulators within the molecular pathways targeted by these therapies, reinforcing its potential for mechanistically grounded treatment stratification.
Open AccessMechanism-aware inference of response to targeted cancer therapies
Targeted therapies like small-molecule inhibitors often work by blocking proteins that cancer cells rely on for survival. Omics based modeling of drug sensitivity alone lack mechanistic grounding. We propose FORGE (Factorization Of Response and Gene Essentiality) a simple yet powerful joint matrix factorization framework that co-models drug response and target gene essentiality, enabling the stratification of promising treatment groups for targeted therapy consideration. FORGE also provides Benefit Score-a predictive score that estimates treatment efficacy from basal gene expression profiles. We validated the predictive performance of FORGE across multiple targeted therapies, including Erlotinib (EGFR inhibitor) and Daporinad (NAMPT inhibitor). Our meta-analysis of large scale in-vitro studies underscores FORGE's ability to identify common determinants of drug vulnerabilities and target gene essentiality. Such convergences were not observed when treatment vulnerabilities and gene essentialities were modeled independently. We also demonstrated the universality of Erlotinib Benefit Scores by transferring transformations learned from high-throughput drug response studies across other published datasets, including the TAHOE-100M single-cell perturbation atlas and patient-derived xenograft studies. FORGE successfully identified key regulators within the molecular pathways targeted by these therapies, reinforcing its potential for mechanistically grounded treatment stratification.
- Open Access
- 15 April 2025
- Published by Microbiology Society in Journal of Medical Microbiology
- Vol. 74 (4) , 002003
Comparative genomic analysis of paired clinical isolates from a patient with recurrent melioidosis reveals a low within-host mutation rate
Introduction. Relapse of melioidosis is not uncommon and can occur due to shorter oral antibiotic therapy in the first episode. In such isolates, low mutation rates were identified amongst paired clinical isolates during relapse, but large-scale structural variants were also common. Hypothesis. Using pair-wise comparison, a low number of mutations, especially amongst the virulence and antibiotic resistance genes, may be present amongst the paired isolates obtained during the study period. Aim. A pair of clinical isolates obtained from a patient with recurrent melioidosis during the study period (January 2018 to June 2021) was analysed for identifying the genomic relatedness and DNA changes that may have caused the relapse. Methodology. Using paired-end Illumina sequencing, following appropriate data quality checks, the genomes were assembled using Shovill pipeline, whilst the variants were called using Snippy. Structural variants were detected using TIDDIT, and functional associations were identified using the STRING database searches. Results. One of the isolates (from the second episode) had a highly fragmented genome, but very few structural variants and SNPs were identified. Both the isolates had similar virulence and antibiotic resistance genes; however, owing to the few structural changes, a slightly lower number of virulence genes were observed. Together, they shared 99.8% of the proteomes, and most variants identified spanned either hypothetical proteins or un-annotated regions. Conclusions. Based on comprehensive genome analysis the two strains were genetically similar, with a few structural variants, implying the second episode to be a relapse rather than a re-infection. There was no difference in the antibiotic resistance or virulence genes that may have explained the relapse.
- 15 April 2025
- Published by Microbiology Society in Journal of Medical Microbiology
- Vol. 74 (4) , 002003
Open AccessComparative genomic analysis of paired clinical isolates from a patient with recurrent melioidosis reveals a low within-host mutation rate
Introduction. Relapse of melioidosis is not uncommon and can occur due to shorter oral antibiotic therapy in the first episode. In such isolates, low mutation rates were identified amongst paired clinical isolates during relapse, but large-scale structural variants were also common. Hypothesis. Using pair-wise comparison, a low number of mutations, especially amongst the virulence and antibiotic resistance genes, may be present amongst the paired isolates obtained during the study period. Aim. A pair of clinical isolates obtained from a patient with recurrent melioidosis during the study period (January 2018 to June 2021) was analysed for identifying the genomic relatedness and DNA changes that may have caused the relapse. Methodology. Using paired-end Illumina sequencing, following appropriate data quality checks, the genomes were assembled using Shovill pipeline, whilst the variants were called using Snippy. Structural variants were detected using TIDDIT, and functional associations were identified using the STRING database searches. Results. One of the isolates (from the second episode) had a highly fragmented genome, but very few structural variants and SNPs were identified. Both the isolates had similar virulence and antibiotic resistance genes; however, owing to the few structural changes, a slightly lower number of virulence genes were observed. Together, they shared 99.8% of the proteomes, and most variants identified spanned either hypothetical proteins or un-annotated regions. Conclusions. Based on comprehensive genome analysis the two strains were genetically similar, with a few structural variants, implying the second episode to be a relapse rather than a re-infection. There was no difference in the antibiotic resistance or virulence genes that may have explained the relapse.
- 28 March 2024
- Published by Elsevier in Indian Journal of Medical Microbiology
- Vol. 48, 100575
Utility of quantitative loop mediated isothermal amplification (qLAMP) assay for the diagnosis of invasive pneumococcal disease (IPD)
- 28 March 2024
- Published by Elsevier in Indian Journal of Medical Microbiology
- Vol. 48, 100575
Utility of quantitative loop mediated isothermal amplification (qLAMP) assay for the diagnosis of invasive pneumococcal disease (IPD)
- Open Access
Comparative pangenome analysis of major pneumococcal genotypes from India
Background Pneumococcal genomes are highly dynamic with varying core genome sizes. The genotype classification system, Global Pneumococcal Sequence Clusters, identified patterns within genotype and antibiotic resistance. Few genotypes like GPSC10 are frequently associated with antimicrobial resistance and high rates of non-vaccine serotypes. Objective To identify and annotate the differences in the core genomes of major GPSC in India, and construct and analyse the Indian Pneumococcal Pangenome (IPPG). Methods Using existing dataset from the Global Pneumococcal Sequencing Project, 618 strains were included. The most frequent GPSCs: GPSC1, GOSC2, GPSC8, GPSC9 and GPSC10 were analyzed separately. Pangenomes were constructed using Panaroo with tuning the family threshold parameter. Differences in protein clusters were identified using Orthovenn3 webserver. Functional annotations were performed by eggNOG, Uniprot and STRING database searches. Results The IPPG core genome size (1615 genes) was similar to those reported previously, with similar distribution of metabolic categories across the five GPSC types. The GPSC10 (1619 genes) and GPSC1 (1909 genes) had the lowest and highest core genome sizes respectively, and these core genomes possessed genes encoding for macrolide and tetracycline resistance. Virulence genes ply, psaA, pce (cbpE), pavA, nanB, lytA, and hysA are detected among all the core genomes. Conclusions There is a genotype specific variation within the core genomes of major GPSCs in India. The presence of antibiotic resistance genes among GPSC1 and GPSC10 core genomes explain widespread drug resistance due to these genotypes. The core virulence genes identified among all the genotypes indicate conserved pathogenesis mechanisms, and can be targets for vaccine development or therapy.
Open AccessComparative pangenome analysis of major pneumococcal genotypes from India
Background Pneumococcal genomes are highly dynamic with varying core genome sizes. The genotype classification system, Global Pneumococcal Sequence Clusters, identified patterns within genotype and antibiotic resistance. Few genotypes like GPSC10 are frequently associated with antimicrobial resistance and high rates of non-vaccine serotypes. Objective To identify and annotate the differences in the core genomes of major GPSC in India, and construct and analyse the Indian Pneumococcal Pangenome (IPPG). Methods Using existing dataset from the Global Pneumococcal Sequencing Project, 618 strains were included. The most frequent GPSCs: GPSC1, GOSC2, GPSC8, GPSC9 and GPSC10 were analyzed separately. Pangenomes were constructed using Panaroo with tuning the family threshold parameter. Differences in protein clusters were identified using Orthovenn3 webserver. Functional annotations were performed by eggNOG, Uniprot and STRING database searches. Results The IPPG core genome size (1615 genes) was similar to those reported previously, with similar distribution of metabolic categories across the five GPSC types. The GPSC10 (1619 genes) and GPSC1 (1909 genes) had the lowest and highest core genome sizes respectively, and these core genomes possessed genes encoding for macrolide and tetracycline resistance. Virulence genes ply, psaA, pce (cbpE), pavA, nanB, lytA, and hysA are detected among all the core genomes. Conclusions There is a genotype specific variation within the core genomes of major GPSCs in India. The presence of antibiotic resistance genes among GPSC1 and GPSC10 core genomes explain widespread drug resistance due to these genotypes. The core virulence genes identified among all the genotypes indicate conserved pathogenesis mechanisms, and can be targets for vaccine development or therapy.
- Open Access
The BinaxNOW antigen detection assay on CSF samples for diagnosis of pneumococcal meningitis
Introduction: The BinaxNOW pneumococcal antigen detection assay is an immuno-chromatographic technique where the cell wall polysaccharide instead of capsular polysaccharide is the target. Although intended for use on urine samples, its role in diagnosis of pneumococcal meningitis has been reported by many others. The current study explores its accuracy when compared with real-time PCR. Methodology: CSF samples from suspected meningitis cases were processed for Gram stain and culture following standard protocols. Antigen was detected by BinaxNOW antigen detection assay for Streptococcus pneumoniae . Previously published primers and protocols were used for real-time PCR targeting autolysin ( lyt A) gene and was considered as the gold standard test. Results: Among the 74 samples tested, pneumococcal antigen was detected in 21.6% (16/74) and qPCR in 11 (14.9%) samples. Antigen detection assay had higher sensitivity (90.9%) when compared with culture (45.5%) while the positive predictive value was the lowest (62.5% and 100% for culture). It was in substantial agreement with the gold standard test with kappa coefficient of 0.685 and an accuracy of 90.5%. Conclusions: The BinaxNOW pneumococcal antigen detection test can be employed for initial screening of samples or point-of-care test owing to its substantial agreement with real-time PCR.
Open AccessThe BinaxNOW antigen detection assay on CSF samples for diagnosis of pneumococcal meningitis
Introduction: The BinaxNOW pneumococcal antigen detection assay is an immuno-chromatographic technique where the cell wall polysaccharide instead of capsular polysaccharide is the target. Although intended for use on urine samples, its role in diagnosis of pneumococcal meningitis has been reported by many others. The current study explores its accuracy when compared with real-time PCR. Methodology: CSF samples from suspected meningitis cases were processed for Gram stain and culture following standard protocols. Antigen was detected by BinaxNOW antigen detection assay for Streptococcus pneumoniae . Previously published primers and protocols were used for real-time PCR targeting autolysin ( lyt A) gene and was considered as the gold standard test. Results: Among the 74 samples tested, pneumococcal antigen was detected in 21.6% (16/74) and qPCR in 11 (14.9%) samples. Antigen detection assay had higher sensitivity (90.9%) when compared with culture (45.5%) while the positive predictive value was the lowest (62.5% and 100% for culture). It was in substantial agreement with the gold standard test with kappa coefficient of 0.685 and an accuracy of 90.5%. Conclusions: The BinaxNOW pneumococcal antigen detection test can be employed for initial screening of samples or point-of-care test owing to its substantial agreement with real-time PCR.
- Open Access
Usefulness of quantitative Loop-Mediated Isothermal Amplification Assay in comparison with real time PCR for the diagnosis of invasive pneumococcal disease
Background Quantitative LAMP (qLAMP) assay is one of the recent and emerging diagnostic tests for infectious diseases. Only a few studies exist comparing this assay with quantitative real-time PCR (qPCR) for the diagnosis of invasive pneumococcal disease (IPD). The present study was performed to compare the diagnostic performance of qLAMP assay with qPCR targeting autolysin gene for the diagnosis of invasive pneumococcal disease. Methodology/Principal Findings Ninety six blood samples and 73 CSF samples from patients clinically suspected with community acquired pneumonia and acute meningitis were tested by qPCR and qLAMP assays using previously published primers and protocols. The qPCR was considered as the gold standard test and the diagnostic performance was assessed by calculating sensitivity, specificity, positive and negative predictive values, and kappa coefficient for the level of agreement between the tests. Chi-squared/Fisher exact test was used to compare categorical variables (positive/negative). Thirty two blood samples and 22 CSF samples were positive by qPCR while 24 and 20 samples were positive by qLAMP assay respectively. The sensitivity of qLAMP assay was only 86.4% and 75% when tested on CSF and blood samples respectively. However, the qLAMP assay was in substantial to almost perfect agreement when compared with qPCR. The results were statistically significant in both sample types (PConclusions The performance of qLAMP assay can vary based on the specimen type. It has very high specificity and had substantial to almost perfect agreement, and thus may be an alternative to qPCR for the diagnosis of IPD.
Open AccessUsefulness of quantitative Loop-Mediated Isothermal Amplification Assay in comparison with real time PCR for the diagnosis of invasive pneumococcal disease
Background Quantitative LAMP (qLAMP) assay is one of the recent and emerging diagnostic tests for infectious diseases. Only a few studies exist comparing this assay with quantitative real-time PCR (qPCR) for the diagnosis of invasive pneumococcal disease (IPD). The present study was performed to compare the diagnostic performance of qLAMP assay with qPCR targeting autolysin gene for the diagnosis of invasive pneumococcal disease. Methodology/Principal Findings Ninety six blood samples and 73 CSF samples from patients clinically suspected with community acquired pneumonia and acute meningitis were tested by qPCR and qLAMP assays using previously published primers and protocols. The qPCR was considered as the gold standard test and the diagnostic performance was assessed by calculating sensitivity, specificity, positive and negative predictive values, and kappa coefficient for the level of agreement between the tests. Chi-squared/Fisher exact test was used to compare categorical variables (positive/negative). Thirty two blood samples and 22 CSF samples were positive by qPCR while 24 and 20 samples were positive by qLAMP assay respectively. The sensitivity of qLAMP assay was only 86.4% and 75% when tested on CSF and blood samples respectively. However, the qLAMP assay was in substantial to almost perfect agreement when compared with qPCR. The results were statistically significant in both sample types (PConclusions The performance of qLAMP assay can vary based on the specimen type. It has very high specificity and had substantial to almost perfect agreement, and thus may be an alternative to qPCR for the diagnosis of IPD.
- Open Access
Structure of Macrolide Efflux Protein (MefA/E) in Streptococcus pneumoniae : An in silico approach
Background: Macrolides are one of the commonest antibiotics used to treat bacterial respiratory tract infections. Resistance to this class of antibiotics is on the rise and is mediated by macrolide efflux (MefA/E) protein as one of the mechanisms. Despite its importance, the structure of this protein is not known yet. Methods: The publicly available MefA/E protein sequence was used to model the structure. Modelling was performed in I-TASSER, and the model was further refined. Its orientation in a membrane was studied using OPM server. Results and conclusions: The structure of MefA/E resembled that of Major Facilitator Superfamily (MFS) proteins, with 13 transmembrane helices. It had a V-shaped conformation, with the wider part towards the outer membrane layer.
Open AccessStructure of Macrolide Efflux Protein (MefA/E) in Streptococcus pneumoniae : An in silico approach
Background: Macrolides are one of the commonest antibiotics used to treat bacterial respiratory tract infections. Resistance to this class of antibiotics is on the rise and is mediated by macrolide efflux (MefA/E) protein as one of the mechanisms. Despite its importance, the structure of this protein is not known yet. Methods: The publicly available MefA/E protein sequence was used to model the structure. Modelling was performed in I-TASSER, and the model was further refined. Its orientation in a membrane was studied using OPM server. Results and conclusions: The structure of MefA/E resembled that of Major Facilitator Superfamily (MFS) proteins, with 13 transmembrane helices. It had a V-shaped conformation, with the wider part towards the outer membrane layer.