SEARCH · Search PubMed

Results for “In silico modeling”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

2,968 recordsLinked to original sources

PubMed2026

In silico identification of DNMT1 inhibitors from the PlantCyc database through computational approach to assess the anti-cancer potential of nutraceutical compounds in breast cancer.

Breast cancer accounts for a disproportionate share of global cancer-related deaths, with 670,000 fatalities and 2.3 million new diagnoses recorded in women during 2022 alone. Existing treatment modalities carry considerable toxicity burdens, and resistance to available agents remains an unresolved clinical problem. DNA methyltransferase 1 (DNMT1), the enzyme chiefly responsible for maintaining genome-wide methylation patterns during DNA replication, has been mapped out as a high-value target in breast cancer because its dysregulation silences tumour suppressor genes through promoter hypermethylation. The present work involves hierarchical in silico workflow to screen 4549 plant-derived compounds from the PlantCyc database (v16.0.3) against the human DNMT1 catalytic domain (PDB ID: 4WXX). Ten top-scoring compounds were taken forward for molecular docking via AutoDock Vina; Quercetin and Kaempferol both recorded the highest binding affinities at -9.5 kcal/mol, Wogonin (-9.3 kcal/mol) and Xanthohumol (-8.1 kcal/mol) also emerged as strong binders. Pharmacokinetic evaluation using ADMET-AI confirmed that all 10 compounds met Lipinski's rule of five, with human intestinal absorption values at or above 0.98. Wogonin and Xanthohumol were selected for a 100 ns all-atom molecular dynamics (MD) simulation in GROMACS due to their well-rounded ADMET profiles and limited existing data on their specific interactions with DNMT1 in breast cancer. Across all measured trajectory metrics, backbone RMSD, residue fluctuation, radius of gyration, solvent-accessible surface area, and intermolecular hydrogen bond count, Wogonin formed a more stable, compact complex. These findings suggest that Wogonin and Xanthohumol are non-toxic nutraceutical candidates suitable for DNMT1 targeted epigenetic therapy, with computational foundation strong enough to facilitate future in vitro and in vivo validation work.

Humans↗

PubMed2026

A versatile reversed-phase liquid chromatography charged aerosol detection method for streamlined monitoring of QS-21 content and stability in liposomal adjuvant formulations.

Identifying and quantifying an active adjuvant along with its degradants in drug formulations is essential for ensuring the safety and efficacy of the drug product. QS-21 is a potent adjuvant that is being evaluated in several clinical trials and is currently formulated in licensed vaccines that protect against shingles, malaria, and RSV. In aqueous environments, QS-21 is subject to hydrolytic degradation that is influenced by pH and temperature, resulting in the formation of a degradant known as QS-21 Hydrolyzed Product, QS-21 HP, which can occur during manufacturing and/or prolonged storage. The intact QS-21 and QS-21 HP induce distinct immune response profiles, making it critical to monitor the degradation of QS-21 in vaccine adjuvant formulations. To date, there has been a paucity of reliable assays for QS-21, its isomers, and degradant QS-21 HP in liposomal adjuvant formulations available that can be transferred seamlessly in quality control (QC) environments. Herein, we introduce a simple and QC-friendly liquid chromatography coupled to a charged aerosol detector (LC-CAD) enabled by stationary phase screening combined with in silico method development optimization. The method exploits 2.7 μm fused-core phenyl hexyl particles, ensuring its versatility in standard and ultra-high pressure LC systems. This approach demonstrates a high correlation between predicted retention time (RT) and experimental outcomes with overall ∆RT < 4%. In addition, this assay shows great linearity, precision, specificity, and accuracy to advance process development characterization of new vaccine formulations.

Liposomes↗

PubMed2026

Epigenetics and In Silico Transcriptome Analysis of Pediatric Acute Myeloid Leukemia.

Pediatric acute myeloid leukemia (AML) is a heterogeneous hematologic malignancy that accounts for about 15%-20% of childhood leukemias. Despite therapeutic advances, relapses remain common, and survival for high-risk patients is below 60%. Unlike adult AML, pediatric AML displays distinct genetic mutations, including FLT3-ITD, NPM1, KMT2A rearrangements, and core-binding factors (CBF) fusions, as well as extensive epigenetic dysregulation. Aberrant DNA methylation, histone modifications, and altered non-coding RNA expressions disrupt hematopoietic differentiation and activate oncogenic transcriptional networks. Recent advances in silico transcriptomic analysis have transformed the study of pediatric AML by integrating gene expression and epigenetic data to identify molecular drivers and regulatory networks. Computational RNA-seq pipelines and pathway analyses have highlighted key epigenetic regulators, including DNMT3A, TET2, and HDACs, as potential therapeutic targets. Multi-omics approaches combining transcriptomic, methylomic, and chromatin accessibility data are increasingly used to define biomarkers for diagnosis, prognosis, and therapeutic response. This review provides a comprehensive overview of the molecular and epigenetic landscape of pediatric AML, emphasizing the power of in silico transcriptome analysis to uncover disease mechanisms, refine patient stratification, and guide the development of precision-based epigenetic therapies aimed at improving long-term outcomes in children with AML.

Humans↗

PubMed2026

Discovery and characterization of multifunctional bioactive peptides from Alaska Pollock (Gadus chalcogrammus) milt: hybrid in silico, in vitro, and proteomic approaches.

The growing demand for multifunctional bioactive peptides has sparked interest in underutilized marine by-products as sustainable bioresources. This study explored Alaska Pollock (Gadus chalcogrammus) milt protein as a novel source of peptides with anti-inflammatory, anti-hypertensive, and anti-diabetic effects. Protein composition was analyzed via LC-MS, followed by in silico digestion and bioactivity prediction. Molecular docking identified peptides targeting DPP-IV, α-glucosidase, ACE, GLP-1 receptor, COX-2, MuRF1, and the 20S proteasome. Among the candidates, a promising peptide (CLPPH) was synthesized and validated in vitro, demonstrating inhibitory effects on nitric oxide production, DPP-IV, ACE, and α-glucosidase. These results highlight CLPPH's potential as a multifunctional bioactive peptide and support the valorization of Alaska Pollock milt as a sustainable source for functional foods and nutraceutical applications.

Animals↗

PubMed2026

Comparative in silico analysis of Apis mellifera immune responses to Varroa destructor and Tropilaelaps mercedesae: Common and mite-specific molecular signatures.

Parasitic mites Varroa destructor and Tropilaelaps mercedesae represent major threats to global honey bee (Apis mellifera) health and productivity, yet comparative molecular insights into host responses remain limited. To address this, we systematically compiled published studies (2015-2025) reporting genes associated with honey bee interactions with V. destructor (11 studies, 87 genes), T. mercedesae (4 studies, 35 genes), and hygienic behavior (6 studies, 44 genes). Gene identifiers were harmonized to the Amel_HAv3.1 genome assembly, yielding three non-redundant sets: 64 Varroa-associated, 34 Tropilaelaps-associated, and 44 hygienic behavior-associated genes. Venn analysis identified 10 overlapping genes (including A0A088A8D5, A0A088ADL8, ABAE_APIME, Def1, Def2, Gapdh, HYTA_APIME, Imd, LOC726783, and Vg), suggesting conserved defense mechanisms, while 41 and 24 genes were uniquely associated with Varroa and Tropilaelaps, respectively. Enrichment analyses revealed Varroa-responsive genes were enriched in immune processes, chitin catabolism, and signaling pathways (Toll/Imd, MAPK, Wnt). Tropilaelaps-associated genes were enriched for antibacterial defense and stress response, with Toll/Imd signaling as the sole significantly enriched pathway. Overlapping genes reinforced core innate immunity activation. Protein-protein interaction network centrality analysis identified key hub genes: Def1, HYTA_APIME, ABAE_APIME, PPO, Imd, PGRP-LC, Vg for Varroa; and ACPH1_APIME, MRJP1, Vg, LOC726783 for Tropilaelaps. Results demonstrate that, despite differences in mite biology, honey bees show a conserved immune response against both parasites, centered on antibacterial defense, humoral immunity, and activation of the Toll/Imd pathway. Although limited by the in-silico nature and research asymmetries reflecting Tropilaelaps' emergence, this curated resource establishes a comprehensive framework for elucidating shared and distinct molecular defense mechanisms. Ultimately, this approach prioritizes diagnostic markers and candidate genes for functional validation and breeding strategies to enhance colony resilience against mite‑driven disease globally.

Animals↗

PubMed2026

A machine learning-derived and functionally validated circadian rhythm signature predicts clinical outcomes and in silico drug sensitivity in colorectal cancer.

BACKGROUND: Colorectal cancer (CRC) displays considerable heterogeneity in clinical outcomes, highlighting the need for reliable prognostic biomarkers. While the aberrant expression of circadian rhythm-related genes has been implicated in cancer pathogenesis, its comprehensive role in CRC progression and predicted therapeutic vulnerabilities remains inadequately characterized. METHODS: Bulk and single-cell RNA-sequencing data were integrated from multiple CRC cohorts. A circadian rhythm signature (CRS) was developed through machine learning algorithms and validated for prognostic value. Comprehensive analyses of tumor microenvironment, genomic alterations, and drug sensitivity were performed. Furthermore, the biological function of the core gene, BHLHE40, was validated in CRC cell lines through CCK-8, EdU, and wound healing assays. RESULTS: Single-cell analysis demonstrated an elevated expression signature of circadian rhythm-related genes in dendritic cells. The optimized CRS, comprising 14 circadian rhythm-related genes, successfully categorized patients into high- and low-risk groups. Patients with a high CRS showed markedly poorer overall survival and computationally inferred immunosuppressive features, including reduced CD8+ T cell infiltration and increased M2 macrophage polarization. Genomic analysis revealed enhanced mutation burden in TP53 and alterations in RTK-RAS/WNT pathways. Notably, in vitro assays confirmed that BHLHE40 is significantly overexpressed in CRC cells. Knockdown of BHLHE40 markedly inhibited tumor cell proliferation and migration. Drug sensitivity profiling identified bexarotene and SMER-3 as potential therapeutic options for high-CRS patients. A nomogram integrating CRS with clinical parameters demonstrated superior predictive accuracy for 1-, 3-, and 5-year survival. CONCLUSIONS: The CRS represents a promising prognostic biomarker that reflects tumor immune status and genomic features, providing valuable insights for personalized treatment strategies in CRC.

Circadian rhythm↗

PubMed2026

Bioactive peptides for meat quality and preservation: Integrating peptidomics and computational screening.

Bioactive peptides generated from meat proteins, fermented meat products, and slaughter by-products have attracted increasing attention as functional molecules for improving meat quality and preservation. In meat systems, peptides can be produced through endogenous postmortem proteolysis, microbial fermentation, gastrointestinal digestion, or controlled enzymatic hydrolysis of underutilized animal by-products. These peptides are closely associated with key meat science endpoints, including postmortem tenderization, oxidative stability, color retention, flavor development, microbial inhibition, and the valorization of processing by-products. However, although high-resolution peptidomics has greatly expanded the identification of meat-derived peptide sequences, their translation into practical meat applications remains limited by matrix interactions, processing stability, sensory constraints, safety concerns, and insufficient validation in real meat systems. This review synthesizes recent advances in meat-related peptidomics and computational screening, including sequence-based prediction, machine learning, molecular docking, molecular dynamics, stability assessment, and safety-oriented filtering. Particular attention is given to how these approaches can prioritize peptides with antioxidant, antimicrobial, flavor-modulating, and preservation-related functions under meat-specific technological constraints. By integrating peptide generation pathways, mass spectrometry-based identification, in silico prioritization, and meat quality endpoints, this review proposes a stage-gated framework for translating meat-derived bioactive peptides from discovery to application. Future research should strengthen matrix-specific validation, standardized peptidomic reporting, and safety assessment to support the use of bioactive peptides in meat quality improvement, clean-label preservation, and circular utilization of meat industry by-products.

Animals↗

PubMed2026

Metagenome-scale modeling to assess microbiome metabolic complementarity for precision microbiota transplantation therapies.

Fecal microbiota transplantation (FMT) holds therapeutic promise beyond recurrent Clostridioides difficile infection, but clinical outcomes remain unpredictable and donor-selection strategies remain limited, in part because the role of donor‒recipient metabolic interactions in shaping the post-FMT community remains poorly understood. Here, we leverage metagenome-scale metabolic modeling to quantify metabolic niche complementarity between donor and recipient microbiomes and predict post-FMT community composition. Using MICOM-derived metabolic models, we show that donor genomes whose metabolic flux profiles are more dissimilar from the recipient community colonize at significantly higher rates in a murine FMT model. In a human IBS trial, the same metric predicted post-FMT community composition via leave-one-out cross-validation and captured known disease-associated alterations in short-chain fatty acid, sulfur, and gas metabolism. We then performed 2,548 in silico FMT simulations between IBS-D/M patients and donors from the OpenBiome biobank to evaluate personalized donor screening, identifying super-donors characterized by high taxonomic diversity, broad metabolic niche coverage, and community interaction networks dominated by cross-feeding rather than competition. Together, these results support metabolic niche complementarity as a potential determinant of post-FMT community composition and provide a mechanistic basis for evaluating donor-recipient metabolic compatibility. This framework offers a scalable approach for generating testable hypotheses for personalized donor selection.

Fecal Microbiota Transplantation↗

PubMed2026

Diversification of yeast proteins as an approach for the development of sustainable food systems.

Despite growing trend in sustainable protein sources, yeast proteins have mainly been explored as a source of bioactive peptides using a monospecies and general protein approach. The contribution of highly abundant protein fractions in the yeast proteome to peptide formation remains insufficiently investigated, limiting a comprehensive understanding of yeast proteins as optimized peptide sources. The current review presents a systematic analysis of yeast proteins as emerging protein sources and evaluates the suitability of high-abundance proteins as bioactive peptide precursors by in silico techniques. Moreover, brewery by-product and single-cell yeast protein approaches are compared in terms of composition and techno-functionality whereas peptide formation mechanisms (in situ and ex situ) and regulatory aspects for food applications are also addressed. Cytoplasmic metabolic proteins, particularly glycolytic enzymes (GAPDH), are identified as highly abundant fractions of the yeast proteome. Proteins associated with cell and organelle membranes also contribute substantially based on cellular localization. These findings imply that such proteins may act as key precursors of yeast-derived bioactive peptides. In silico hydrolysis with Alcalase suggests a tendency toward the generation of short-chain peptides (3-11/14 aa), which may support biological activity. Moreover, peptide profiles appear to vary across yeast species, highlighting the role of species diversity in peptide generation. While single-cell yeast protein allows more controlled production than brewery by-products, nucleic acid content in both may limit applications. Overall, yeast proteins appear to be metabolically adaptable and species-diverse sources for various biological peptides.

Saccharomyces cerevisiae↗

PubMed2026

Integrated exome and mitochondrial genome sequencing reveals the genetic landscape of primary mitochondrial diseases: findings from a large Tunisian cohort.

Primary mitochondrial diseases are a heterogeneous group of neurometabolic disorders recognized as the most common metabolic genetic diseases. They manifest at any age, affecting any tissue or organ, especially those with high energy demands, and are caused by pathogenic variants in both mitochondrial and nuclear genomes. Here, we aimed to describe the genetic spectrum of a Tunisian pediatric cohort with suspected mitochondrial diseases. We recruited 47 unrelated families who underwent exome sequencing as a first-tier test followed by whole mitochondrial genome sequencing for unsolved cases. Dedicated bioinformatic pipelines and prediction tools were used to determine the potential disease-causing variants. Sanger sequencing confirmed the presence and segregation within parents. For the newly identified variants, structural modeling was conducted to study the impact of these variants on protein structure and motions. Dual genome sequencing yielded a molecular diagnosis in 33/47 families (70%) and 18/47 (38%) showed disease-causing variants in genes encoding mitochondrial proteins. Among them, four families disclosed novel variants in FASTKD2, SERAC1 and GATB, which were supported by in-depth in silico and structural analyses demonstrating their deleterious effect. The remaining families (32%, 15/47) disclosed other metabolic and neurological disorders. An exome-first strategy delivers a high diagnostic yield in Tunisia, where consanguinity remains high and simultaneously captures mitochondrial and non-mitochondrial etiologies. Mitochondrial sequencing remains indispensable in the case of an inconclusive exome. Thus, our data expand the clinical and genetic spectrum of primary mitochondrial diseases in Tunisia, an underrepresented and admixed population.

Humans↗

PubMed2026

Circular RNAs in amyotrophic lateral sclerosis.

Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disorder characterized by the progressive loss of motor neurons, with most cases lacking a clear genetic basis. Emerging evidence highlights the involvement of non-coding RNAs, particularly circular RNAs (circRNAs), in disease onset and progression. Here, we investigated circRNAs implicated in ALS and related motor neuron diseases (MNDs). Here, we provide a general overview of circular RNA metabolism and cellular functions. We then present our systematic literature review that identified ALS-associated circRNAs, followed by in silico analyses of 15 circular RNA candidates that were selected based on the most compelling data regarding ALS. Our results revealed that several circular RNAs regulate ALS-related genes, such as unfolded protein response, oxidative stress, cell cycle regulation, and apoptosis. Protein-RNA interaction analysis further showed that ALS-related circRNAs can sponge 20 RNA-binding proteins. Additionally, molecular docking analysis demonstrated that ALS-associated FUS variants significantly alter its binding affinity to circular RNAs. RNA-seq data from ALS patients confirmed significant alterations in the expression of host genes of ALS-related circRNAs and hub proteins in ALS-affected CNS tissues. Collectively, our findings identify circRNAs as potential key contributors to ALS pathogenesis.

Amyotrophic Lateral Sclerosis↗

PubMed2026

Uncovering hidden complexity in the Apis mellifera mitotranscriptome: a polyadenylation-centered perspective.

Mitochondrial transcription is gaining increasing attention as researchers seek to better understand the full coding potential of mitochondrial DNA (mtDNA). Emerging evidence suggests that mtDNA may encode additional elements beyond classical oxidative phosphorylation genes, pointing to a more complex transcriptional architecture than previously recognized. In this study, we explored the mitochondrial transcriptome of Apis mellifera (Insecta: Hymenoptera), with a particular focus on polyadenylation-associated features. Our analysis revealed that both sense and antisense transcripts undergo polyadenylation, although transcript abundance and poly(A) tail lengths varied markedly across mitochondrial genes. Several transcripts exhibited alternative isoforms, either extended or truncated, frequently including intergenic regions. These regions may represent functional non-coding elements or structural variants rather than conventional untranslated regions (UTRs). Interestingly, some transcripts also contained non-templated nucleotide additions particularly cytosine residues immediately upstream of the poly(A) tails. Monocistronic units that included portions of downstream intergenic regions were among the most abundantly represented, suggesting a possible regulatory role for these sequences. To experimentally validate our in silico findings, we performed RT-qPCR to assess relative gene expression and applied 3' RACE-PCR to define transcript boundaries. These approaches confirmed the presence of multiple transcript isoforms and supported the involvement of polyadenylation in shaping mitochondrial RNA diversity. Together, our findings reveal a previously underappreciated level of complexity in the A. mellifera mitochondrial transcriptome and highlight the potential regulatory significance of polyadenylation dynamics and intergenic region transcription.

Animals↗

PubMed2026

Olive leaf protein hydrolysates yield gastro-resistant peptides with antioxidant and anti-inflammatory potential: peptidomics, in vitro validation and molecular docking analyses.

Olive (Olea europaea L.) leaves are an abundant olive-oil by-product and a promising feedstock for sustainable valorisation. An olive leaf protein isolate (OLPI) from olive-leaf powder (OLP) was enzymatically hydrolysed to yield seven hydrolysates (OLPHs). All showed notable antioxidant activity as whole hydrolysate matrices (EC₅₀ = 0.11-0.28 mg mL-1); likely reflecting the combined contribution of released peptides and co-extracted phenolic compounds; the 15-min Alcalase product (OLPH15A) showed high activity with the shortest processing time. Its INFOGEST digest (dOLPH15A) attenuated LPS-induced inflammation in Caco-2 cells, down-regulating pro-inflammatory and up-regulating anti-inflammatory genes. Peptidomics identified 7037 peptides in OLPH15A and 534 in dOLPH15A, from which twenty gastro-resistant sequences were prioritised for in silico analysis. Multi-tool prediction and docking highlighted four peptides, GAAGGIGQPL, QSAYPGTGPL, GGGAGGGDGGIL and LDAQFPGVN, with favourable predicted affinity for the TLR4/MD2 complex, suggesting that they may contribute to the observed immunomodulatory response. These findings position olive leaves as a viable source of protein hydrolysate-based ingredients with antioxidant and anti-inflammatory potential, advancing the valorisation of olive-oil by-products.

Olea↗

PubMed2026

The gonadal matrisome and its correlation with sex change in the ricefield eel Monopterus albus.

The matrisome is a comprehensive list of genes in the genome of an organism, which encodes proteins constituting or interacting with the extracellular matrix (ECM). The gonadal ECM is important for folliculogenesis and spermatogenesis. This study characterized the composition of the matrisome and the expression of matrisome genes in the gonad of ricefield eel, a protogynous sex-changing teleost, during sex change. A total of 838 matrisome genes were identified in the genome of ricefield eel through an in-silico orthology-based approach, of which 482, 443, 429, and 570 matrisome genes were shown to be expressed in the gonads of female (F), early intersexual (EI), mid-intersexual (MI), and late intersexual (LI) fish, respectively. Differentially expressed matrisome genes (DEMGs) were observed across all the sexual stages as well as in each category of ECM components. Analysis of DEMGs in the comparison between EI and F revealed dramatic upregulation of adam8a, mmp9, and s100a11 while downregulation of col4a5, col15a1b, clec3ba, f13a1, and ccl44, which were further confirmed by qPCR analysis. Together, these findings revealed significant changes in the expression of many matrisome genes, particularly three regulator genes, adam8a, mmp9 and f13a1, as female ricefield eels initiate sex change, suggesting that gonadal tissues undergo dramatic remodeling involving the regression of ovarian tissues and the development of testicular tissues to facilitate this process. These data provide valuable resources for further unraveling the roles of matrisome genes in gonadal development of ricefield eel and other vertebrates.

Animals↗

PubMed2026

Accurate quantification of canine mitochondrial DNA copy number from canine blood and brain samples.

Acute brain injury is difficult to evaluate in veterinary medicine and tools to investigate the potential involvement of mitochondrial involvement are limited. The brain is highly enriched in mitochondria and contains thousands of copies of mitochondrial DNA (mtDNA) per cell, but robust methods for quantifying mitochondrial DNA copy number (mtDNA-CN) in canine tissues are lacking. We describe the development of a quantitative real-time PCR assay for absolute measurement of mtDNA-CN which was validated in canine blood and brain tissue. To minimize amplification of nuclear mitochondrial insertion sequences (NumtS) and repetitive regions, species-specific oligonucleotide primers were designed following in silico genomic filtering. The assay was applied to a small pilot cohort comprising blood samples from dogs with and without acute brain injury (n = 4-6 per group) and cerebral cortex samples (n = 1 per group) to assess feasibility and biological plausibility. In non-brain injury dogs, blood mtDNA-CN ranged from 98 to 288 copies per nuclear genome (mean 193 ± 72), while values in brain-injured cases ranged from 163 to 228 copies per genome (mean 200 ± 33). Cerebral cortex samples exhibited higher mtDNA-CN than blood, consistent with known tissue-specific mitochondrial enrichment. In a single brain-injured case with serial sampling, mtDNA-CN increased over five days. This study presents a validated assay and pilot data for mtDNA-CN quantification in canine samples. While not powered for biomarker evaluation, this method may enable future studies of mitochondrial dynamics in canine brain injury and metabolic disease.

Animals↗

PubMed2026

Statistical test to compare the linkage model and the admixture model based on central limit results.

In the Admixture Model, the probability that an individual carries a certain allele at a specific marker depends on the allele frequencies in K ancestral populations and the proportion of the individual's genome originating from these populations. The markers are assumed to be independent. The Linkage Model is a Hidden Markov Model that extends the Admixture Model by incorporating linkage between neighboring loci. We prove consistency and asymptotic normality of maximum likelihood estimators for the ancestry of individuals in the Linkage Model, complementing earlier results by (Pfaff et al., 2004; Pfaffelhuber and Rohde, 2022; Heinzel, 2025) for the Admixture Model. These results are used to prove that a statistical test that allows for model selection between the Admixture Model and the Linkage Model is an asymptotic level-α-test. Finally, we demonstrate the practical relevance of our results by applying the test to real-world data from The 1000 Genomes Project Consortium (2015).

Genetic Linkage↗

PubMed2026

Modelling the effects of biological intervention in a dynamical gene network.

Cellular response to environmental and internal signals can be modeled by dynamical gene regulatory networks (GRN). In the literature, three main classes of gene network models can be distinguished: (1) non-quantitative (or data-based) models which do not describe the probability distribution of gene expressions; (2) quantitative models which fully describe the probability distribution of all genes co-expression; and (3) mechanistic models which allow for a causal interpretation of gene interactions. We propose two rigorous frameworks to model gene alteration in a dynamical GRN, depending on whether the network model is quantitative or mechanistic. We explain how these models can be used for design of experiment, or, if additional alteration data are available, for validation purposes or to improve the parameter estimation of the original model. We apply these methods to the Gaussian graphical model, which is quantitative but non-mechanistic, and to mechanistic models of Bayesian networks and penalized linear regression.

Gene Regulatory Networks↗

PubMed2026

The landscape of pruning for large language models: A systematic review and unified taxonomy.

Confronting the inherent tension between the exceptional capabilities and the immense computational costs of Large Language Models (LLMs), pruning has become a crucial technique for achieving efficient deployment. However, a systematic analytical framework dedicated specifically to LLM pruning remains absent. In this paper, we aim to bridge this gap. We first elucidate the theoretical foundations that underpin the effectiveness of pruning, namely overparameterization and redundancy, and then propose a multidimensional taxonomy that organizes existing approaches along the axes of granularity, timing, and criteria. Building upon this unified perspective, we further analyze performance recovery mechanisms and the broader evaluation ecosystem, while also exploring forward-looking challenges such as interpretability, automation, and hardware-algorithm co-design. Through this comprehensive synthesis, we seek to provide an integrated and coherent analytical lens for advancing both research and practice in LLM pruning.

Large Language Models↗