Table 1_rMAP-Candida: a modular Dockerized WDL/Cromwell workflow for reproducible Candida species typing, assembly-contiguity assessment, antifungal-resistance marker screening, and phylogenomic surveillance.xlsx

Background

Candida spp. infections are an increasing public health concern, particularly in settings where laboratory mycology, genomic surveillance infrastructure, and antifungal susceptibility testing remain limited. Accurate species identification, reproducible assembly assessment, conservative genomic screening for antifungal-resistance markers, and interpretable phylogenomic outputs are essential for surveillance and outbreak preparedness. However, fungal whole-genome sequencing workflows remain fragmented, difficult to reproduce across computing environments, and insufficiently adapted for implementation in low-resource public health genomics settings.

Methods

We developed rMAP-Candida, a modular, Dockerized WDL/Cromwell workflow for paired-end Candida spp. whole-genome sequencing analysis. The workflow performs read quality control and trimming with fastp, Candida-focused species typing using Kraken2/Bracken, de novo assembly with MEGAHIT, assembly-contiguity assessment with QUAST, optional genome-completeness assessment using Compleasm or BUSCO, antifungal-resistance marker screening using ChroQueTas/FungAMR-derived outputs, species-aware core-SNP phylogenomics, pairwise SNP-distance summarization, closest-neighbor analysis, and integrated HTML surveillance reporting. To improve independent reproducibility, the repository includes a quick-start local Cromwell test, a corrected two-sample input JSON, documented checks for public container and database access, and a two-sample reproducibility run.

Results

rMAP-Candida generated reproducible species assignments, assembly-contiguity metrics, optional completeness summaries, antifungal-resistance marker outputs, species-aware phylogenomic summaries, pairwise SNP-distance tables, closest-neighbor summaries, and integrated HTML reports. In the Ugandan validation dataset, the workflow identified six principal species groups, dominated by Candida albicans, followed by Candida tropicalis, Pichia kudriavzevii, Nakaseomyces glabratus, Clavispora lusitaniae, and Candida parapsilosis. The integrated report summarized 24 antifungal-resistance marker hits and a median assembly N50 of 35,946 bp. Species-aware phylogenomics was performed for eligible species groups, while ineligible or skipped groups were explicitly reported with reasons. The report also distinguished “no curated genomic antifungal-resistance marker detected” from phenotypic susceptibility, supporting conservative interpretation of resistance-screening outputs.

Conclusion

rMAP-Candida provides a portable, reproducible, modular, and surveillance-oriented WDL/Cromwell workflow for Candida spp. genomic analysis. By integrating species identification, assembly-contiguity assessment, optional completeness evaluation, antifungal-resistance marker screening, species-aware phylogenomics, SNP-distance summarization, closest-neighbor reporting, and HTML reporting, the workflow supports training, research, and applied fungal genomic surveillance in low-resource and other implementation settings.