Biology subjects

Klempir, O.

Publications and source records attributed to Klempir, O..

5 recordsLinked to original sources

bioRxiv2026

An end-to-end in silico framework for truth-based validation and regression testing of clinical genome sequencing pipelines, from simulated reads to reportable findings

Clinical genome sequencing pipelines change often, and each change requires proportionate revalidation. Guidelines accept in silico data as a supplement but leave variant interpretation outside pipeline validation, and, to our knowledge, no framework tests a pipeline against a known truth up to the reportable finding, so report specificity is not measured. We present a source-available framework that tests a pipeline end to end, from sequencing reads to the report, against a truth fixed before any run. It plants expert-curated pathogenic and benign variants from the same genes, and optionally synthetic variants that no database holds, into the GIAB HG002 germline background and records the expected class and report decision of each variant. It then simulates reads and runs secondary and tertiary analysis as black boxes through exchangeable adapters. Findings are compared with the truth at the call-set and report levels, each failure is assigned to the first pipeline stage where it appears, and a variant-by-variant comparison of two runs serves as a regression test. We applied the framework to nf-core/sarek with DeepVariant and to the ACMG/AMP classifier BIAS-2015, using 243 ClinGen-curated variants in five test sets. Single-nucleotide variant recall against the HG002 benchmark was 99.5% on simulated and 99.3% on real reads, and every planted variant was present in the reads and called with the correct genotype. The report included 115 of 117 pathogenic and likely pathogenic variants (98.3%) but withheld only 106 of 126 variants that must not be reported (84.1%), mainly because variants of uncertain significance were upgraded and recessive carriers were reported. Without ClinVar-derived criteria, report sensitivity fell to 74.4%. Most false reports of one test set came from diploid calling of the male chromosome X, and haploid calling removed them without side effects. In supplementary experiments, indels were called completely up to 20 bp, and BIAS-2015 left 74% of synthetic loss-of-function variants unreported because it combined criteria by the 2015 rules. The framework measures report-level sensitivity and specificity and localizes regressions, so it can supplement validation on human samples and support change control.

bioinformatics↗

bioRxiv2026

Analyzing Genomic Foundation Models for Viral Sequence Identification

Rapid and accurate identification of viral sequences underpins clinical diagnostics, epidemiological surveillance, and the safety testing of biological products, yet the established alignment-based methods such as BLAST are inherently closed-set, with accuracy tied to how well a query is already represented in the reference database. Recent advances in genomic foundation models (GFMs) have enabled alignment-free approaches to sequence classification. However, their performance relative to traditional methods remains incompletely characterised. We evaluated GFMs, specifically the Nucleotide Transformer (NT) and DNABERT families, for viral sequence classification using a dataset derived from the Reference Viral Database (RVDB). Sequence embeddings were generated with mean, max, and CLS pooling, classified by nearest-neighbour search in an embedding index and compared with BLAST as a reference. Matching BLAST provided a particularly stringent benchmark here, because as a closed-set method it compares each query directly against the exact reference sequences it stores, whereas a foundation model must rely on representations shaped by broad, general-purpose pre-training. On sequences of standard length, the best model, NTv3-650M, reached an accuracy of 0.93, approaching the 0.97 achieved by BLAST. On long genomic sequences, evaluated on a much smaller test set, the difference narrowed further, with NTv3-650M reaching 0.95 against 0.98 for BLAST. Robustness analyses showed that all GFMs were markedly more sensitive to base masking than BLAST. Both methods indexed the same labelled reference sequences, and BLAST compared queries with them directly at the nucleotide level, so approaching it from frozen, general-purpose representations without any parameter update is a meaningful result. NTv3-650M emerged as the most promising candidate for further optimisation and deployment in bioinformatics applications.

bioinformatics↗

bioRxiv2021

BioCompute Objects to communicate a viral detection pipeline with potential for use in a regulatory environment

The volume of nucleic acid sequence data has exploded in recent years, and with it, the challenge of finding and transforming relevant data into meaningful information. Processing the abundance of data can require a dynamic ecosystem of customized tools. As analysis pipelines become more complex, there is an increased difficulty in communicating analysis details in a way that is understandable yet of sufficient detail to make informed decisions about results or repeat the analysis. This may be of particular interest to institutions and private companies that need to communicate complex computations in a regulatory environment. To meet this need for standard reporting, the open source BioCompute framework was developed as a standardized mechanism for communicating the details of an analysis in a concise and organized way, and other tools and interfaces were subsequently developed according to the standard. The goal of BioCompute is to streamline the process of communicating computational analyses. Reports that conform to the BioCompute standard are called BioCompute Objects (BCOs). Here, a comprehensive suite of BCOs is presented, representing interconnected elements of a computation that is modeled after those that might be found in a regulatory submission, but which can be shared publicly. Because BCOs are human and machine readable, they can be displayed in customized ways to further improve their utility, and an example of a collapsible format is shown. The work presented here serves as a real world implementation that imitates actual submissions, providing concrete examples. As an example, a pipeline designed to identify viral contaminants in biological manufacturing, such as for vaccines, is developed and rigorously tested to establish a rate of false positive detection, and is described in a BCO report. That pipeline relies on a specially curated database for alignment, and a set of synthetic reads for testing, both of which are also descriptively packaged in their own BCOs. All of the sufficiently complex processes associated with this analysis are therefore represented as BCOs that can be cross-referenced, demonstrating the modularity of BCOs, their ability to organize tremendous complexity, and their use in a lifelike regulatory environment.

bioinformatics↗

bioRxiv2020

Mutation Maker, An Open Source Oligo Design Software for Mutagenesis and De Novo Gene Synthesis Experiments

Protein engineering is the discipline of developing useful proteins for applications in research, therapeutic and industrial processes by modification of naturally occurring proteins or by invention of de novo proteins. Modern protein engineering relies on the ability to rapidly generate and screen diverse libraries of mutant proteins. However, design of mutant libraries is typically hampered by scale and complexity, necessitating development of advanced automation and optimization tools that can improve efficiency and accuracy. At present, automated library design tools are functionally limited or not freely available. To address these issues, we developed Mutation Maker, an open source mutagenic oligo design software for large-scale protein engineering experiments. Mutation Maker is not only specifically tailored to multi-site random and directed mutagenesis protocols, but also pioneers bespoke mutagenic oligo design for de novo gene synthesis workflows. Enabled by a novel bundle of orchestrated heuristics, optimization, constraint-satisfaction and backtracking algorithms, Mutation Maker offers a versatile toolbox for gene diversification design at industrial scale. Supported by in-silico simulations and compelling experimental validation data, Mutation Maker oligos produce diverse gene libraries at high success rates irrespective of genes or vectors used. Finally, Mutation Maker was created as an extensible platform on the notion that directed evolution techniques will continue to evolve and revolutionize current and future-oriented applications.

bioinformatics↗

bioRxiv2018

A Deep Learning Genome-Mining Strategy Improves Biosynthetic Gene Cluster Prediction

Natural products represent a rich reservoir of small molecule drug candidates utilized as antimicrobial drugs, anticancer therapies, and immunomodulatory agents. These molecules are microbial secondary metabolites synthesized by co-localized genes termed Biosynthetic Gene Clusters (BGCs). The increase in full microbial genomes and similar resources has led to development of BGC prediction algorithms, although their precision and ability to identify novel BGC classes could be improved. Here we present a deep learning strategy (DeepBGC) that offers more accurate BGC identification and an improved ability to extrapolate and identify novel BGC classes compared to existing tools. We supplemented this with downstream random forest classifiers that accurately predicted BGC product classes and potential chemical activity. Application of DeepBGC to bacterial genomes uncovered previously undetectable BGCs that may code for natural products with novel biologic activities. The improved accuracy and classification ability of DeepBGC represents a significant step forward for in-silico BGC identification.

bioinformatics↗