Biology subjects

Morang'a, C. M.

Publications and source records attributed to Morang'a, C. M..

6 recordsLinked to original sources

bioRxiv2026

The Reasoning Paradox in Scientific Sandboxes: Why Extended Reasoning Fails in Autonomous Bioinformatics and How to Fix It.

Frontier large language models (LLMs) trained with extended reasoning, such as DeepSeek-R1 and OpenAI o1, have established state-of-the-art benchmarks in formal logic, competitive programming, and mathematics. In interactive, empirical scientific sandboxes, however, their efficacy remains counter-intuitive. In the Bioinformatics Benchmark (BixBench), extended reasoning models paradoxically degraded in multi-turn environments, leading benchmark authors to exclude them from primary agentic evaluations. In this paper, we conduct the first systematic forensic audit of the internal cognitive monologue ( ... ) of open-weight reasoning models (DeepSeek-R1-Distill-Qwen-14B) and matched instruction-tuned controls (Qwen2.5-Coder-14B) deployed inside containerized computational biology sandboxes on Nvidia H100 GPUs. Our forensic analysis uncovers two distinct systemic failure modes: (1) The PubMed Prior Trap, in which reasoning models engage in epistemic rationalization within their cognitive stream, falsely convincing themselves that empirical datasets are missing and substituting empirical computation with speculative literature recall (e.g., estimating adjusted p-values based on "typical experimental outcomes"); and (2) Tag Language Nesting Collisions, in which models conflate agent-level XML action tags with Python syntax, producing terminal interpreter syntax errors. To resolve this paradox, we introduce Dual Track Decoupled Reasoning (DTDR), an agent scaffolding architecture that evicts internal cognitive monologues from multi-turn history, injects strict empirical grounding guardrails, and dynamically intercepts runtime tool invocations. On a stratified diagnostic benchmark spanning single-cell RNA-seq, bulk differential expression (DESeq2), and genomic variant prioritization (n=22 per arm), DTDR rescues the 14B reasoning model from 0.0% to 9.09% accuracy (2/22), doubling the performance of the matched non-reasoning coder control (4.55%, 1/22) while slashing code execution errors by 81% (13.6% vs. 72.7%) and reducing average steps to solution by 66% (1.77 vs. 5.18 steps). All replication pipelines, prompts, and trajectory manifests are openly released.

bioinformatics↗

bioRxiv2026

Tensorizing Pairwise Structural Analysis: GPU-Accelerated RMSD and FCC Clustering in HADDOCK3 Without Structural Regression.

Integrative macromolecular docking pipelines generate thousands of candidate structural decoys, relying on downstream pairwise ensemble analysis, including Root-Mean-Square Deviation (RMSD) matrices and Fraction of Common Contacts (FCC) clustering, to identify biologically relevant binding poses. However, pairwise structural comparisons scale quadratically (O(N^2)), and canonical workflows in HADDOCK3 historically rely on disk-bound trajectory serialization and iterative CPU set operations, creating a quadratic computational and storage bottleneck for large ensembles. Here, we demonstrate that these operations can be reformulated as continuous linear algebra on GPU-resident tensors: structural superposition via batched Singular Value Decomposition (SVD) and residue contact overlap via binary matrix multiplication (M = A A^T). Across empirical benchmarks spanning N = 200 to 10,000 models (evaluating up to 49,995,000 decoy pairs), our GPU implementation achieves a 6.41-fold compute speedup on Apple Silicon (MPS) and up to 12.2-fold end-to-end acceleration in an N = 5,000 ablation on NVIDIA CUDA by eliminating trajectory disk I/O. Crucially, the implementation achieves zero downstream structural regression relative to canonical baselines, preserving identical cluster assignments (Adjusted Rand Index ARI = 1.0000, zero differing assignments) with sub-microangstrom RMSD concordance (Delta_max <= 4.07 x 10^-7 Angstrom on single-precision MPS; machine-precision Delta_max <= 8.88 x 10^-16 Angstrom on double-precision CUDA). The accelerated code, test suites, and benchmarking workflows are openly accessible on GitHub under the Apache-2.0 license.

bioinformatics↗

bioRxiv2026

Characterising recent antimalarial resistance in West Africa: Insights from amplicon sequencing of 17,384 Plasmodium falciparum infection samples

Plasmodium falciparum (P. falciparum) infection remains a significant public health threat in West Africa, where chemoprevention and first-line therapies are key interventions against malaria. However, the development and spread of resistance to commonly used antimalarials poses a growing threat to the efficacy of these strategies. This study characterises the recent landscape of antimalarial resistance in West Africa by analysing targeted amplicon sequences from 17,384 P. falciparum infection samples. Across countries, the prevalence of the pyrimethamine resistance-associated dhfr triple mutant allele (51I/59R/108N) exceeded 80%, while its combination with the sulphadoxine resistance-associated dhps 437G exceeded 60% of infections. Unlike the parasite genotypes in East Africa, the prevalence of the dhps 540E mutant was low (1.5%), whereas dhps 436A was common (43.8%). The chloroquine resistance marker crt 76T showed greatest geographic heterogeneity, ranging from low prevalence in Ghana (1.3%) to very common in The Gambia (64.9%). Non-synonymous mutants of kelch13 were uncommon, most with unknown relevance to artemisinin resistance and observed for the first time in Africa. However, mutants that are artemisinin resistance-associated elsewhere were detected in three infection samples from Ghana (574L, 561H, 469Y), and one in Cameroon (538V). This large-scale genomic surveillance of P. falciparum infections highlights the need for ongoing monitoring of drug resistance and for data integration throughout the region.

genomics↗

bioRxiv2025

Discovery of Novel Trypanothione Reductase Inhibitors through Pharmacophore Modeling, virtual Screening and Molecular Dynamics Simulations. New insights for Human African Trypanosomiasis

Human African Trypanosomiasis (HAT), caused by Trypanosome brucei, remains a critical health concern, with treatments such as Melarsoprol hampered by toxicity and resistance. To address this, we aimed to identify inhibitors targeting Trypanothione Reductase (T.b TR), a key enzyme in the parasites survival. A pharmacophore model was derived from Melarsoprol and validated using Receiver Operating Characteristic (ROC) curves and Enrichment Factor (EF). Virtual screening of compound libraries identified two promising candidates, VS-1 and VS-2, based on superior docking scores. Subsequent molecular dynamics simulations confirmed the stability of the ligand-enzyme complexes, while binding free energy calculations revealed strong binding affinities for VS-1 and VS-2. Per-residue decomposition pointed critical interactions with active site residues, including MET 260, ASN 130, and HIS 128, contributing to the compounds stability and activity. These findings suggest that VS-1 and VS-2 are potential inhibitors of T.b TR, with better efficacy than Melarsoprol. However, further in vitro studies, including IC50 determination, cell cycle analysis, and morphological assays, are essential to confirm the therapeutic potential of these compounds for the treatment of HAT.

bioinformatics↗

bioRxiv2022

Nanopore sequencing for real-time genomic surveillance of Plasmodium falciparum

Malaria is a global public health priority causing over 600,000 deaths annually, mostly young children living in Sub-Saharan Africa. Molecular surveillance can provide key information for malaria control, such as the prevalence and distribution of antimalarial drug resistance. However, genome sequencing capacity in endemic countries can be limited. Here, we have implemented an end-to-end workflow for P. falciparum genomic surveillance in Ghana using Oxford Nanopore Technologies, targeting antimalarial resistance markers and the leading vaccine antigen circumsporozoite protein (csp). The workflow was rapid, robust, accurate, affordable and straightforward to implement, and could be deployed using readily collected dried blood spot samples. We found that P. falciparum parasites in Ghana had become largely susceptible to chloroquine, with persistent sulfadoxine-pyrimethamine (SP) resistance, and no evidence of artemisinin resistance. Multiple Single Nucleotide Polymorphism (SNP) differences from the vaccine csp sequence were identified, though their significance is uncertain. This study demonstrates the potential utility and feasibility of malaria genomic surveillance in endemic settings using Nanopore sequencing.

genomics↗

bioRxiv2022

Molecular epidemiology of Infectious Spleen and Kidney Necrosis Virus (ISKNV) in Ghanaian cultured tilapia

1Infectious Spleen and Kidney Necrosis Virus (ISKNV) is increasingly gaining more attention globally, due to its highly significant economic impact on the aquaculture industry. In late 2018, unusually high levels of mortality (60-90%) was reported in some intensive tilapia cage culture systems in Ghana. Preliminary investigations confirmed the involvement of ISKNV, a viral pathogen noted for fatal systemic infections in many fish species. As a follow-up on the outbreak situation, and post-mass vaccination of affected fish farms, the need to investigate further the molecular epidemiology and phylogeography of the virus across Lake Volta became paramount. In this study, a multiplexed PCR assay and MinION nanopore sequencing of the Major Capsid Protein (MCP) were performed to investigate the presence and genotype of ISKNV in tilapia collected from 30 randomly selected farms spread across Lake Volta. Fish with and without clinical signs were included in the molecular detection of the virus from brain, kidney and spleen tissues. ISKNV was detected at 80% prevalence with fry and juvenile fish being most affected. Phylogenetic analysis of the MCP revealed that all 35 isolates from 14 different farms were ISKNV genotype I with near- 100% homology to the 2018 outbreak strain. Vaccination and heat shock treatment; the main specific interventions currently employed to control the viral pathogen have not achieved much success and ISKNV remains a threat to the growth of the aquaculture industry in Ghana. The outcome of this study can be useful in improving fish health management and biosecurity policies in the aquaculture industry.

molecular biology↗