Building a machine-learning taxon classifier for genomic classification in malaria mosquitoes

This project aims to build a machine-learning taxon classifier for genomic classification in malaria mosquitoes, operating directly on raw sequencing reads (FASTQ files). Accurate taxonomic identification is essential for malaria vector genomics, since many Anopheles mosquitoes are morphologically indistinguishable and misclassification can lead to incorrect downstream analyses, flawed epidemiological interpretation, and suboptimal control strategies. Current approaches typically rely on genotype calls produced by variant-calling pipelines. While effective, these approaches introduce additional computational cost and complexity, and they are less convenient in settings where only raw reads are available. The goal of this project is to develop a lightweight FASTQ-based classifier that can guide a sample to the correct major taxonomic group or downstream genomic resource without requiring full genotyping or fine-scale species splitting. Rather than attempting detailed resolution within closely related complexes at the outset, the classifier will focus on robust assignment to broader taxonomic groupings or reference frameworks, such as An. gambiae, An. funestus, or An. stephensi.