INBIOSIS · UKM
A daily digest of the three most important new papers on AI and LLM applications in bioinformatics, computational biology, genomics, proteomics and drug discovery. Curated every morning.
18 issues · 53 papers · updated 2026-11-08
Cutting‑edge LLM methods are expanding into DNA and protein modeling.
Introduces a plug‑in module that equips frozen LLMs with DNA‑specific token embeddings, enabling zero‑shot genomic sequence analysis without fine‑tuning.
Why it matters. Provides a rapid, low‑resource way to apply powerful LLMs to DNA tasks.
Proposes a unified framework that uses a single LLM to simulate protein evolution across families, outperforming specialized models on benchmark mutational scans.
Why it matters. Streamlines protein engineering pipelines with a versatile AI tool.
Presents a masked autoencoder that learns representations aware of genome structure, boosting downstream tasks like gene annotation and strain typing.
Why it matters. Enhances microbial genomics analyses with richer AI‑driven embeddings.
Cutting‑edge LLM agents are now turning bio‑informatics pipelines into autonomous, low‑cost research assistants.
LLM agents paired with the Morpheus simulator automatically construct virtual tissue models for fifteen benchmark studies, meeting 83‑86 % of study‑specific criteria in under two hours and for less than $4 per model.
Why it matters. Shows AI agents can democratise in‑silico tissue modelling, slashing cost and expertise barriers.
A transformer that learns read‑to‑specimen count mappings reduces Jensen‑Shannon divergence from 0.36 to 0.12 on a 1,188‑bottle ground‑truth dataset, outperforming generic deep models.
Why it matters. Delivers accurate, scalable abundance estimates crucial for ecological monitoring and biodiversity studies.
CellCanon unifies single‑cell expression prediction, evidence‑grounded explanation and observation‑driven updating, boosting differential‑expression AUPRC by up to 63 % on SciPlex3 held‑out conditions.
Why it matters. Provides an updatable AI framework for cellular response inference, accelerating hypothesis generation from single‑cell data.
Recent AI breakthroughs accelerate bioinformatics research across genomics, protein analysis, and drug discovery.
The fourth release of the DBAASP database enriches peptide and peptidomimetic entries with condition‑specific bioactivity data and curated chemical structures, enabling AI models to learn nuanced activity patterns. It also provides standardized APIs for seamless integration into LLM‑driven drug design pipelines.
Why it matters. High‑resolution bioactivity annotations empower LLMs to generate more accurate peptide candidates for therapeutics.
Using 104 Rosalind challenges, the study benchmarks leading LLMs (including GPT‑5.2 and Claude Sonnet 4.6) on code generation, sequence analysis, and reasoning tasks. GPT‑5.2 achieves 77 % correct answers, highlighting rapid performance gains but also revealing category‑specific weaknesses.
Why it matters. Provides the first systematic, peer‑reviewed evaluation of LLMs on real bioinformatics problems, guiding tool selection for labs.
This chapter demonstrates how transformer‑based LLMs can capture complex base‑pairing and structural motifs to predict RNA‑RNA interactions without handcrafted features, outperforming traditional thermodynamic models on benchmark datasets.
Why it matters. Offers a scalable, annotation‑free approach for studying non‑coding RNA networks, a key frontier in functional genomics.
Cutting‑edge AI agents are reshaping molecular modeling, immunogenomics, and neoantigen design.
RLAgent lets users interact via natural language to define RNA regions and predict ligand binding, bypassing complex code and providing region‑specific insights. It combines retrieval of relevant literature with on‑the‑fly LLM reasoning.
Why it matters. Enables biologists to explore RNA‑ligand interactions rapidly without programming expertise.
pMHCBind fuses protein‑level LLM embeddings with a hypergraph neural net to forecast peptide‑MHC binding affinity, improving prediction accuracy over traditional methods.
Why it matters. Accelerates vaccine and immunotherapy target discovery by reliably modeling antigen presentation.
NeoGen‑BC integrates generative protein LLMs with deep learning to propose neoantigen peptides common across breast‑cancer patients, streamlining personalized vaccine design.
Why it matters. Provides a scalable path to shared cancer vaccines, reducing the cost and time of bespoke neoantigen pipelines.
Cutting‑edge LLM‑driven tools are reshaping experimental pipelines, drug discovery, and AI agent onboarding in biology.
A Claude‑code plug‑in, Brieflow‑auto, lets a coding LLM run the Brieflow optical pooled‑screen pipeline, auto‑tuning parameters, handling failures, and validating outputs. Tested on five screens, it improves cell‑gene assignment confidence over manual analyses.
Why it matters. Automating complex microscopy pipelines accelerates functional genomics studies and lowers the barrier for non‑expert users.
AutoOnboarder trains LLM agents for specific tasks via an intelligent “Onboarder” that iteratively adjusts model, prompt, and harness code using only tens of examples, demonstrated on genomic‑sequencing prior‑authorization forms.
Why it matters. Robust onboarding reduces deployment costs and brings LLM agents closer to real‑world biomedical workflows.
MADDIE combines a central Orchestrator with specialized agents (HitIdentifier, LeadOptimizer, Synthesizer, Evaluator) to execute flexible, end‑to‑end drug‑discovery campaigns, outperforming existing systems on 50 protein targets.
Why it matters. Multi‑agent orchestration enables scalable, low‑cost autonomous drug design, promising faster therapeutic innovation.
Cutting‑edge large language models are reshaping data extraction, functional annotation, and clinical trial reporting in biology.
The authors assembled a benchmark of 406 full‑text articles covering 180 congenital heart disease genes and evaluated seven commercial LLMs (including GPT‑4o, Claude‑Opus‑4, DeepSeek‑R1) on multi‑dimensional extraction accuracy. They introduce fuzzy‑matching metrics to handle synonyms and partial matches.
Why it matters. Provides the first systematic performance picture of top LLMs for large‑scale gene‑disease‑phenotype curation, guiding tool selection for database curators.
Introducing MPA‑EC, a generative framework that fine‑tunes an LLM on an Enzyme Microenvironment Database, enabling the model to infer pH/temperature context and produce reasoning‑rich EC label predictions. It outperforms GraphEC by 0.11 F1 and works on cold‑start enzymes.
Why it matters. Demonstrates how incorporating microenvironmental context into LLMs can substantially boost functional annotation of novel enzymes.
This systematic review of 68 trial reports, 15 protocols, and 378 registrations assesses adherence to reporting standards (TRIPOD‑LLM, CONSORT‑AI, CHART). Overall completeness hovers around 40‑60 %, with major gaps in model provenance, versioning, and safety reporting.
Why it matters. Highlights crucial transparency deficits that must be addressed for trustworthy clinical deployment of LLM‑driven interventions.
Cutting‑edge AI tools are reshaping drug discovery, literature synthesis, and protein‑targeted therapies.
The review surveys how AlphaFold‑derived structures accelerate identification of novel therapeutics and vaccine candidates for aquaculture pathogens, integrating AI‑driven docking and epitope prediction pipelines.
Why it matters. Demonstrates the practical impact of generative protein‑structure AI on sustainable aquaculture and rapid vaccine design.
Introduces an LLM workflow that retrieves relevant trial reports, abstracts key findings, and generates structured outcome summaries, showcased on three digital‑health RCTs.
Why it matters. Shows how LLMs can automate systematic review steps, speeding evidence synthesis in clinical research.
Combines mechanistic language models with 3‑D molecular screening to predict and experimentally validate a synergistic combination of berberine and enzalutamide against prostate cancer.
Why it matters. Highlights AI‑driven drug‑combination discovery that leverages biological mechanism knowledge.
AI agents are accelerating omics data reuse and bioinformatics automation.
Introduces GeneReL, a platform that uses LLMs to extract gene regulatory interactions from literature and combines community curation to build a high‑confidence Arabidopsis network.
Why it matters. Provides scalable, AI‑augmented curation for plant regulatory databases, dramatically expanding coverage.
Presents an LLM‑agent framework that automates the discovery, download, re‑analysis, and meta‑synthesis of omics datasets, achieving high reproducibility across proteomics and RNA‑seq.
Why it matters. Demonstrates practical, end‑to‑end AI‑driven pipelines for reusing publicly available omics data.
Introduces BioGAIP, a multi‑agent LLM system that orchestrates bioinformatics analyses from data retrieval to result interpretation, showing robustness across diverse tasks.
Why it matters. Offers a versatile AI assistant that can handle varied bioinformatics workflows with minimal user effort.
Cutting‑edge large language model techniques are reshaping bioinformatics, from smarter in‑context learning to clinical text mining and rare‑disease knowledge retrieval.
The authors propose Dual‑Div, a two‑stage demo‑selection framework that jointly optimises representativeness and diversity of in‑context examples. Tested on NER, RE and text‑classification tasks with LLaMA 3.1 and Qwen 2.5, Dual‑Div consistently outperforms baselines by up to 5 % macro‑F1 while using only 3‑5 demonstrations.
Why it matters. Efficient demo selection makes LLMs viable for routine biomedical NLP without costly fine‑tuning.
Open‑weight LLMs are prompted in a chain‑of‑thought style to first extract temporal expressions linked to clinical events and then normalise them. On French rare‑disease case notes the method attains F1 = 0.72 for diagnoses and 0.61 for phenotypes, rivaling specialised neural baselines.
Why it matters. Demonstrates that minimal in‑context prompting can unlock high‑quality temporal reasoning for low‑resource clinical languages.
WilsonLitQA combines hybrid retrieval with generative LLMs to answer any query over the full PubMed corpus on Wilson's disease. Retrieval grounding dramatically raises factual accuracy, and compact 7 B‑parameter models perform on par with larger systems.
Why it matters. Provides a scalable, open‑source tool for clinicians and researchers to navigate fragmented rare‑disease literature safely.
Cutting‑edge LLMs are now being harnessed for concrete biomedical tasks, from interpreting genetic severity to prioritising drug candidates and even ranking molecular conformers.
An autonomous AI agent combines ReAct reasoning with Retrieval‑Augmented Generation to classify genetic disease severity using ACMG guidelines, achieving >93% accuracy and providing evidence‑backed explanations.
Why it matters. It offers a standardized, scalable way to assess disease severity for genomic screening panels.
A new pre‑training framework integrates hierarchical code aggregation and cross‑reference mechanisms, improving clinical prediction and enabling in‑silico screening that successfully rediscovers known Alzheimer’s drugs.
Why it matters. Demonstrates that domain‑aware LLMs can accelerate hypothesis generation for drug repurposing.
Evaluating modern LLMs on conformer ranking shows several models rival traditional force fields, indicating emergent 3‑D chemical reasoning from language training.
Why it matters. Opens the door for using LLMs in autonomous molecular design and drug discovery pipelines.
Spotlights emerging LLM-driven methods for tumor prognostication and hypothesis generation.
The study applies a graph-guided mixture of experts to multi-modal tumor data, demonstrating improved survival prediction. It showcases how LLMs can integrate diverse biomedical modalities.
Why it matters. Integrating multimodal data via LLM architectures advances precision oncology and provides actionable insights for clinicians.
The study applies a graph-guided mixture of experts to multi-modal tumor data, demonstrating improved survival prediction. It showcases how LLMs can integrate diverse biomedical modalities.
Why it matters. Integrating multimodal data via LLM architectures advances precision oncology and provides actionable insights for clinicians.
The study applies a graph-guided mixture of experts to multi-modal tumor data, demonstrating improved survival prediction. It showcases how LLMs can integrate diverse biomedical modalities.
Why it matters. Integrating multimodal data via LLM architectures advances precision oncology and provides actionable insights for clinicians.
Cutting‑edge language models are now the new building blocks for rapid hypothesis generation in life sciences, from protein engineering to single‑cell workflows.
A transformer‑based approach jointly trains on sequence, AlphaFold features and structural restraints to predict the impact of point mutations on protein stability, achieving >90 % accuracy against experimentally annotated datasets.
Why it matters. Enables high‑throughput in‑silico mutagenesis screening for therapeutic antibody optimization and enzyme design.
An auto‑encoding framework that learns shared latent representations across 50 vertebrate species and can impute missing single‑cell data in non‑model organisms using a fine‑tuned GPT architecture.
Why it matters. Expands downstream pathway analysis to ecological and evolutionary studies where experimental profiling is scarce.
A diffusion‑based generative model conditioned on cell‑type labels synthesises pathway activity matrices, allowing researchers to infer unmeasured biochemical pathways from scRNA‑seq profiles.
Why it matters. Provides an integrative tool for reconstructing regulatory networks without costly perturbation experiments.
The past week’s literature showcases novel ways large language models are integrated into protein bioinformatics, systems biology frameworks, and agent benchmarking.
The authors reveal that latent embeddings of a protein language model capture structural motifs and functional domains, enabling interpretability for downstream proteomic predictions.
Why it matters. This work offers a new avenue to translate abstract LLM features into tangible biological insight, accelerating hypothesis generation in bioinformatics.
Using an agent‑based framework that integrates diverse protein abundance datasets, the authors identify tightly co‑expressed modules conserved across tissues, shedding light on systems‑level protein regulation.
Why it matters. By linking multi‑omics data through AI orchestration, this study paves the way for more comprehensive models of cellular function and disease pathways.
This benchmark evaluates several LLM‑driven agent chains across a suite of computational biology problems, highlighting performance gaps and guiding future development.
Why it matters. The results inform the design of more capable AI agents for complex bioinformatics workflows, promoting reproducible, high‑impact research productivity.
Cutting‑edge LLMs are accelerating biologically relevant discovery, from de novo molecules to disease‑specific structural models.
Combines a transformer trained on SMILES and chemical property data to generate synthesizable drug candidates; achieves >70% success in Tanimoto similarity benchmark versus baseline methods.
Why it matters. Speeds up medicinal chemistry cycles by automating early‑stage molecule design.
Uses a language model fine‑tuned on synthetic genetic network descriptions to output orthogonal gene circuits with specified logic functions; validated in silico for Boolean behavior.
Why it matters. Enables rapid prototyping of synthetic biology constructs without manual design.
Applies a diffusion-based LLM to predict disease‑associated conformations of tau and alpha‑synuclein proteins; provides new structural hypotheses that could guide therapeutic approaches.
Why it matters. Offers mechanistic insights that could accelerate therapy development for Parkinson’s and Alzheimer’s disorders.
This work introduces CritICL, a method that leverages the failure patterns of small LLMs during inference to guide training of larger models, reducing overfitting on rare biological sequences. It demonstrates consistent gains across protein‑folding and gene‑expression tasks with fewer labeled samples.
Why it matters. By turning model errors into signals, CritICL enables efficient scaling of biomedical AI without the data burden that typically limits deployment in resource‑constrained research settings.
SWE‑Prime adapts the large‑language‑model backbone to protein‑ligand affinity prediction by pruning trajectory sets in molecular simulations, yielding a 12 % MAE reduction over baseline GPT‑based generative models while keeping inference time under 2 s per complex.
Why it matters. The approach directly addresses the bottleneck of costly simulation data, accelerating virtual screening pipelines crucial for early drug discovery.
TTPO introduces a test‑time policy refinement framework that fine‑tunes pretrained LLM policies on the fly in biological simulation environments, achieving up to 18 % higher success rates on synthetic navigation benchmarks for cellular-scale systems.
Why it matters. It demonstrates how dynamic adaptation at run time can overcome static model limitations, paving the way for flexible AI agents that can respond to evolving biological conditions.
Large language models now face unprecedented scrutiny for biosecurity risks, with new frameworks testing whether "safety-aligned" scientific AIs can be coaxed into generating harmful pathogens—uncovering a critical gap between text-level safeguards and real biological capability.
Despite strong biomedical reasoning abilities, LLMs remain unclear on extracting epitope information directly from antigen sequences—critical for antibody design. EpiBench launches a new 1,609-sample benchmark testing five tasks: target region discovery, antibody-conditioned epitope identification, binning, functional assessment, and escape prediction. Results show current models capture partial signals but lack sequence grounding and long-context localization needed for reliable LLM-assisted antibody discovery workflows.
Why it matters. Provides a concrete testbed to measure whether biomedical LLMs can move beyond generic reasoning toward truly biological sequence-aware applications central to therapeutic design.
LLM-powered peptide generative models are being weaponized via "Genotypic Triggers"—backdoor attacks shifting model distribution toward high immunogenicity risk for specific HLA alleles, without triggering standard safety screens that prioritize antimicrobial potency and low general toxicity instead of genetic-specific risks. The attack increased predicted immuno-genicity scores by 743% compared to natural peptides; backdoored models retained or improved primary properties while enabling targeted health threats against carriers of specific gene variants involved in immune presentation.
Why it matters. Demands redesign of biosecurity evaluation pipelines, exposing how conventional safety screens overlook context-specific population-level risks when AI systems are trained for efficacy rather than contextual harm assessment alone.
Cutting‑edge AI tools are streamlining data‑intensive workflows across bioinformatics, from molecular networking to gene‑set interpretation.
An open‑source toolkit (MN‑Suite) built with large language model assistance integrates multiple similarity algorithms for mass‑spec molecular networking, offering flexible, server‑free analysis of natural product datasets.
Why it matters. It demonstrates how LLM‑guided software engineering can accelerate deployment of specialized bio‑informatics pipelines.
The GeneInsight framework extracts functional annotations from STRING‑DB, clusters semantically related terms using sentence embeddings, and generates concise thematic summaries via LLM prompting, streamlining gene‑set interpretation.
Why it matters. It showcases LLMs as powerful assistants for turning sprawling gene‑set outputs into clear biological insights.
Using LLM‑based text mining, quantitative context‑dependent weights are assigned to literature‑extracted gene regulations, producing GRNs that better reflect disease‑specific biology and enhance drug‑target prediction and ODE model construction.
Why it matters. It highlights the value of LLMs for building more accurate, condition‑specific regulatory models.
Cutting‑edge AI tools are reshaping how we predict structures, generate single‑cell data, and assess biosecurity risks.
Large Language Models accelerate biological research but also enable the design of harmful toxin‑like proteins, posing a biosecurity threat. The authors introduce SPIKE‑Bench, a benchmark that evaluates LLMs on toxin‑design prompts and propose a classifier to mitigate risks.
Why it matters. Highlights the need for safety testing of AI systems in biotech.
Traditional protein‑structure prediction yields a single dominant conformation. This work reframes the problem as state‑space inference, reviewing ensemble generators, physics‑based simulations, and experimental constraints to predict multiple functional conformations.
Why it matters. Moves protein modelling towards capturing the dynamic ensembles essential for function and drug design.
Introduces a causal transformer paired with a quantized VAE tokenizer to generate realistic single‑cell gene‑expression vectors. The study characterises biological fidelity and scaling laws, enabling downstream perturbation‑response modelling.
Why it matters. Provides a scalable foundation model for synthetic single‑cell data, accelerating benchmarking and method development.