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

Recent Submissions

  • Thumbnail Image
    Item type:Item, Access status: Open Access ,
    New Additions of Books - September 2026
    (2026-09) Srinivasa Ramanujan Library
  • Thumbnail Image
    Item type:Item, Access status: Open Access ,
    Monthly Bulletin of New Publications - September 2026
    (2026-09) Srinivasa Ramanujan Library
  • Thumbnail Image
    Item type:Item, Access status: Open Access ,
    Teaching Language Models to Forecast Research Success Through Comparative Idea Evaluation
    (Association for Computational Linguistics, 2026-07) MULE, SRUJAN P.; Garikaparthi, Aniketh; Patwardhan, Manasi; Dept. of Data Science
    As language models accelerate scientific research by automating hypothesis generation and implementation, a new bottleneck emerges: evaluating and filtering hundreds of AI-generated ideas without exhaustive experimentation. We ask whether LMs can learn to forecast the empirical success of research ideas before any experiments are run. We study comparative empirical forecasting: given a benchmark-specific research goal and two candidate ideas, predict which will achieve better benchmark performance. We construct a dataset of 11,488 idea pairs grounded in objective outcomes from PapersWithCode. While off-the-shelf 8B-parameter models struggle (30% acc.), SFT dramatically boosts performance to 77.1%, outperforming GPT-5 (61.1%). By framing evaluation as a reasoning task via Reinforcement Learning with Verifiable Rewards (RLVR), we train models to discover latent reasoning paths, achieving 71.35% acc. with interpretable justifications. Through additional ablations and out-of-distribution tests, we show robustness to surface-level heuristics and transfer to both a cross-domain time-split test set and an independently constructed test set. Our results demonstrate that compute-efficient small language models can serve as effective, objective verifiers, offering a scalable path for autonomous scientific discovery.
  • Thumbnail Image
    Item type:Item, Access status: Metadata only ,
    Multifaceted effects of N-glycosylation on amyloidogenic κ light chains in AL amyloidosis
    (Elsevier B.V., 2026-06) PURI, SARITA; Valentina Speranzini et al.; Dept. of Biology
    Light chain (LC) amyloidosis (AL) is a fatal disorder caused by extracellular aggregation of monoclonal immunoglobulin LCs. While both λ and κ isotypes can be involved, κ-LCs account for only ∼20% of cases, and their aggregation mechanisms remain less understood. Recent evidence suggests that N-glycosylation influences κ-LC aggregation and is strongly linked to AL amyloidosis. To investigate this, we examined patient-derived κ-LCs in both glycosylated and unglycosylated forms. Mass spectrometry confirmed the presence of complex-type N-glycans. Biophysical analyses showed that glycosylation enhances structural compactness, stabilizes the native structure, and promotes cooperative unfolding. Glycosylated κ-LCs also exhibited reduced conformational dynamics. Functionally, N-glycosylation significantly improved LC secretion efficiency and extracellular stability. These findings suggest that N-glycosylation acts as a protective factor against amyloid formation and enhances LC accumulation outside the cell. Overall, our study highlights the multifaceted and context-dependent role of N-glycosylation in modulating κ-LC behavior, with important implications in AL pathogenesis.
  • Thumbnail Image
    Item type:Item, Access status: Metadata only ,
    Wunen(s) help navigate Primordial Germ Cells by attenuating Hedgehog signaling
    (eLife Sciences Publications Ltd., 2026-07) ROY, AMRITA; ROY, ADHEENA ELSA; Ibragimov, Airat; DaSilva, Juliana; KUMAR, KUNDAN; Schedl, Paul; KAMAT, SIDDHESH S.; RATNAPARKHI, GIRISH S.; DESHPANDE, GIRISH; Dept. of Biology
    Directed cell migration is a vital process that depends on the combined activities of attractive and repulsive cues. As it is essential for normal development, the precise identity of guidance signals and the underlying molecular and cellular mechanisms is being rigorously investigated. In a Drosophila embryo, PGC migration is orchestrated by non-cell autonomous repulsive and attractive cues, controlled by Wunen(s) - Wunen and Wunen2 and, HMGCoA-reductase (Hmgcr), respectively. Hedgehog (Hh), a PGC attractant, is potentiated by Hmgcr. We demonstrate that Wunen(s) employ both nonautonomous and autonomous modes to inhibit Hh signaling. Consistently, in embryos maternally compromised for wunen, mesodermal cells and PGCs accumulate excess Hh, leading to precocious clumping of the PGCs. This behaviour is reminiscent of PGC-specific loss of patched (ptc) – the Hh receptor and an antagonist of Smoothened (Smo), a G protein-coupled receptor (GPCR), involved in Hh signal transduction. Consistently, Wunen(s) inhibit membrane localization of Smo. Conversely, simultaneous overexpression of wunen mitigates PGC scattering induced by ectopic hmgcr expression. Finally, unbiased lipidomics of embryonic extracts after maternal knockdown of wunen confirms disruptions in lipid metabolism. We discuss the mechanistic underpinnings of Wunen(s) involvement in repressing Hh signalling to engineer PGC migration.