Comprehensive evaluation of haplotype phasing tools with different strategies across diverse sequence technologies

Haplotype phasing is a computational technique used to determine the allelic arrangement in a chromosome from genotype data. This process is crucial for understanding genetic diversity and inheritance patterns, which are essential for fields like genetic research, personalized medicine, and population genetic studies. In this study, we conducted a comprehensive evaluation of five state-of-the-art haplotype phasing tools under different strategies including statistical-based Beagle5, Eagle2 and Shapeit5, and read-based WhatsHap, and HapCut2, across various sequencing platforms including Illumina, PacBio, and ONT. Our findings indicate that statistical-based tools generally provide better phasing continuity and are less dependent on the sequencing technology used, whereas read-based tools offer superior performance with long-read datasets but struggle with short reads due to insufficient sequence length. In addition, statistical-based tools were found to be less resource-intensive compared to their read-based counterparts. The insights gained from this benchmarking provide a valuable guide for researchers in selecting the most appropriate phasing tools, potentially enhancing the accuracy and efficiency of genetic analysis and supporting advancements in genetic research methodologies.