IISER Pune Digital Repository

A repository of scholarly research output from the Indian Institute of Science Education and Research (IISER) Pune.

Browse publications, theses, and other research output produced by the IISER Pune community.

Email: libraryservices@iiserpune.ac.in

Communities in Digital Repository

Select a community to browse its collections.

Now showing 1 - 3 of 3

OTHER RESOURCES 605
PUBLICATIONS & PATENTS 7249
THESES & PROJECT REPORTS 3006
Theses submitted to IISER Pune in partial fulfilment of the requirements of the BS-MS dual degree, MSc. and Ph.D programmes. Also, project reports submitted by the students.

Recent Submissions

Access status: Metadata only ,
Integrating Radiomics and Radiogenomics to Advance Precision Cancer Diagnostics and Treatment
(Wiley, 2026-07-31) Vinodini, D.; KULKARNI, RISHABH; Kundal, Kavita; Akula, Jyothiraditya Sai Ram; Kumar, Rahul; Dept. of Biology; Gautam, Vibhav_Ed.; Kumar, Brijesh_Ed.; Singh, Surya Pratap_Ed.; Hussain, Nazar_Ed.
Radiogenomics has emerged as a transformative interdisciplinary field that integrates quantitative imaging, genomics, artificial intelligence (AI), and multi-omics data to advance precision oncology. The extraction of quantitative imaging biomarkers from modalities such as computed tomography (CT), magnetic resonance imaging (MRI), positron emission tomography (PET), magnetic resonance spectroscopy (MRS), and digital pathology assists in the characterization of tumor phenotype, intratumoral heterogeneity, and microenvironmental alterations beyond conventional visual interpretations, which is enabled by radiomics. Integrating the genomic, transcriptomic, proteomic, metabolomic, and clinical data-derived features with imaging provides a non-invasive framework for understanding molecular alterations, disease progression, treatment response, and recurrence patterns. Radiogenomics analysis improved significantly with the artificial intelligence approaches, which include machine learning, deep learning, transfer learning, graph neural networks, and explainable AI by enabling automated feature extraction, multimodal data integration, and predictive modeling. They have also demonstrated substantial clinical relevance in glioblastoma, non-small cell lung cancer, breast cancer, prostate cancer, liver cancer, and brain metastases. Furthermore, the combination of radiomics with the liquid biopsy technologies like circulating tumor DNA and circulating tumor cells, has increased the monitoring of the non-invasive tumor and the dynamic assessment of treatment response. The AI-driven clinical decision support system that is built upon the radiogenomics frameworks is enhancing and supporting risk stratification, adaptive therapy planning, and personalized therapeutic interventions. There are significant barriers, even though there is considerable progress, that continue to limit large-scale clinical implementation, including data heterogeneity, lack of standardization, annotation variability, limited external validation, interpretability concerns, privacy preservation, and evolving regulatory requirements. The developing fields like federated learning, foundation models, digital twins, and real-time AI-enabled imaging workflows are expected to improve these challenges and the scalability, robustness, and clinical translation of radiogenomic systems. Together, the integration of radiomics, radiogenomics, and AIdriven image-omics by enabling comprehensive, non-invasive, and data-driven approaches for cancer diagnosis, prognostication, treatment planning, and longitudinal disease monitoring redefines precision oncology.
Access status: Metadata only ,
Single‐Cell Transcriptomics
(Wiley, 2026-07-31) Kar, Sushree Sangita; Chakraborty, Tuhin; Dutta, Suvranil; DILEEP, K. AKASH; Sengupta, Avik; Kumar, Rahul; Dept. of Biology; Gautam, Vibhav_Ed.; Kumar, Brijesh_Ed.; Singh, Surya Pratap_Ed.; Hussain, Nazar_Ed.
The field of oncology has undergone significant change enabled by single-cell transcriptomics, allowing more precise analysis of tumor ecosystems which had previously been inaccessible using bulk profiling. This chapter outlines the steps involved in the transition from bulk RNA sequencing (bulk RNA-seq) to single-cell methodologies, detailing the evolution of experimental workflows and the computational frameworks that support them. Single-cell transcriptomics applications in oncology are discussed, particularly the way single-cell RNA sequencing (scRNA-seq) elucidates intratumoral heterogeneity and the complexity of the tumor microenvironment (TME), including its neoplastic, immune, and stromal constituents. These findings have, in turn, facilitated the discovery of predictive and prognostic biomarkers, treatment resistance mechanisms, and applicable therapeutic targets. The chapter will highlight advanced applications in precision medicine, therapeutic selection, and drug discovery. We detailed the gap of direct clinical application and highlighted the integration of multi-omics, artificial intelligence (AI)/machine learning (ML), and spatial transcriptomics (ST) as emerging future avenues. Single-cell methodologies promise more precise and personalized oncology by linking cellular heterogeneity to patient outcomes, representing a fundamental shift in the field.
Access status: Metadata only ,
A finer view of the parameterized landscape of labeled graph contractions
(Elsevier B.V., 2026-02) Mathur, Yashaswini; TALE, PRAFULLKUMAR; Dept. of Mathematics
We study the Labeled Contractibility problem, where the input consists of two vertex-labeled graphs G and H, and the goal is to determine whether H can be obtained from G via a sequence of edge contractions. Lafond and Marchand (2025) [42] initiated the parameterized complexity study of this problem, showing it to be 𝖶⁡[1]-hard when parameterized by the number k of allowed contractions. They also proved that the problem is fixed-parameter tractable when parameterized by the treewidth 𝗍𝗐 of G, via an application of Courcelle's theorem with a non-elementary dependence on the parameter. In this work, we present a constructive fixed-parameter algorithm for Labeled Contractibility with running time 2𝒪⁡(𝗍𝗐2) ⋅|𝑉⁡(𝐺)|𝒪⁡(1). We also prove that unless the Exponential Time Hypothesis (ETH) fails, it does not admit an algorithm running in time 2𝑜⁡(𝗍𝗐2) ⋅|𝑉⁡(𝐺)|𝒪⁡(1). This result adds Labeled Contractibility to a small list of problems for which a 2Θ⁡(𝗍𝗐2) dependence is optimal under ETH. We further strengthen existing hardness results by showing that the problem remains NP-hard even when both input graphs have bounded maximum degree. We also investigate parameterizations by (𝑘+ 𝛿⁡(𝐺)), where 𝛿⁡(𝐺) denotes the degeneracy of G, and rule out the existence of subexponential-time algorithms. This answers a question on subexponential fixed-parameter tractability raised by Lafond and Marchand (2025) [42]. We additionally provide an improved FPT algorithm running in time (𝛿⁡(𝐻)+1)𝑘 ⋅|𝑉⁡(𝐺)|𝒪⁡(1). Finally, we analyze a brute-force algorithm for Labeled Contractibility with running time |𝑉⁡(𝐻)|𝒪⁡(|𝑉⁡(𝐺)|), and show that this running time is optimal under ETH.
Access status: Open Access ,
HEAL Webzine - September 2026, Issue 01
(Dr. SHALINI SHARMA, 2026-09-15) SHARMA, SHALINI; Jamwal, Nidhi
HEAL is grounds-up, bilingual, collaborative, solutions journalism in action; emerging from CARE, an action research project at IISER Pune. It brings together grassroot reporters, researchers, scientist, practitioners and artists in a continuum of care contributing to constructive climate journalism.
Access status: Metadata only ,
Rudra Kavi Virachit Rashtraoudhavanshmahakavya
(Shri Shivsamarth Seva Prakashan, 2026-09) PATEKAR, DEEPAK_Ed.; Bandivadekar Pelapkar, Sai_Trans.
NA