Biology subjects

Papciak, J.

Publications and source records attributed to Papciak, J..

3 recordsLinked to original sources

bioRxiv2026

BIOTIA-DX RESISTANCE Achieved the Best Antimicrobial Resistance Phenotype Prediction Accuracy at CAMDA 2026

We present BIOTIA-DX RESISTANCE (BDXR), our submission to the CAMDA 2026 AMR Challenge. This work extends our CAMDA 2025 submission [1] to a new set of six species-drug pairs and adds k-mer-based feature engineering (both targeted and whole-genome) for pairs where the 2025 gene-presence base model underperforms. BDXR achieved a mean accuracy of 86.1% across the six pairs on the CAMDA 2026 test set, ranking first on four pairs, tied for first on Streptococcus pneumoniae (penicillin), and second on Campylobacter jejuni (nalidixic acid); per-pair test accuracy ranged from 69.9% (C. jejuni, nalidixic acid) to 98.8% (S. pneumoniae, penicillin). We refer the reader to our 2025 preprint [1] for the underlying workflow, dataset curation, and clinical motivation; this preprint focuses on the results and methodological changes that are new in 2026.

microbiology↗

bioRxiv2025

BIOTIA-DX RESISTANCE Achieved the Best Antimicrobial Resistance Phenotype Prediction Accuracy at CAMDA 2025

We have developed BIOTIA-DX RESISTANCE (BDXR), a bioinformatic tool for predicting antimicrobial resistance (AMR) from whole genome sequencing of microbial isolates. BDXR achieved the best accuracy of any submission to the CAMDA 2025 AMR Challenge. This years challenge focused on predicting AMR phenotype which is a more complex problem than the detection of AMR marker genes, the focus of some prior years. BDXR achieved an overall F1 score of 89% on the training set and 84.1% on the challenge test set. Accuracy varied across the 9 species and drug pairs in the competition from an F1 score of 98.4% (Campylobacter jejuni, tetracycline) to 78.5% (Pseudomonas aeruginosa, cefatzidime). BDXR is based on curation of global datasets, machine learning-based predictions from input data, and highly stringent prepreprocessing of input data and databases.

microbiology↗

bioRxiv2024

Forecasting Dengue Fever in Brazil Using Multimodal Data, Including Climate Data

Dengue fever is a major tropical disease transmitted by Aedes mosquitoes. Dengue affects more than 120 countries with highly variable year to year infection rates. Despite high variability, dengue has a clear relationship to climate factors and human demography. Global trends to higher temperatures and greater disorderly urban development are increasing the scale and scope of dengue risk. Dengue has complex human immunity with 4 known serotypes that make multiple infections possible. Accurate forecasting of dengue fever would allow for appropriate interventions and improved public health outcomes. We demonstrate, GeoSeeq Dengue, a forecasting model for dengue fever in Brazil. GeoSeeq Dengue predicts dengue outbreaks monthly in 5,570 Brazilian municipalities at 1, 3, and 6 months ahead of the outbreak. Model accuracy compares favorably to a historical baseline model, making it a promising model for informing public health response. We evaluate how different types of input variables effect model accuracy and explore how this model could be adapted to other countries. This model could inform public health responses to dengue including targeting vector control programs, public health education messaging, and the newly launched dengue vaccine rollout.

bioinformatics↗