FastGA: Fast Genome Alignment

FastGA finds alignments between two genome sequences more than an order of magnitude faster than previous methods that have comparable sensitivity. Its speed is due to (a) a carefully engineered architecture involving only cache-coherent MSD radix sorts and merges, (b) a novel algorithm for finding adaptive seed hits in a linear merge of sorted k-mer tables, and (c) a variant of the Myers adaptive wave algorithm [1] to find alignments around a chain of seed hits that detects alignments with up to 25-30% variation. It further does not require pre-masking of repetitive sequence, and stores millions of alignments in a fraction of the space of a conventional CIGAR-string [2] using a trace-point encoding that is further compressed by the ONEcode data system [3] introduced here. As an example, two bat genomes of size 2.2Gbp and 2.5Gbp can be compared in a little over 2 minutes using 8 threads on an Apple M4 Max laptop using 5.7GB of memory and producing 1.05 million alignments totaling 1.63Gbp of aligned sequence that cover about 60% of each genome. The output "ALN"-formatted file occupies 66MB. This file can be converted to a PAF file with CIGAR strings in 6 seconds, where the PAF representation is a significantly larger 1.03GB file. FastGA is freely available at github: http://www.github.com/thegenemyers/FASTGA along with utilities for viewing inputs, intermediate files, and outputs and transforming outputs into other common formats. Specifically, FastGA can, in addition to its highly efficient ONEcode representation, output PSL-formatted alignments, or PAF-formatted alignments with or without CIGAR strings explicitly encoding the alignments. There is also a utility to chain FastGAs alignments and display them in a dot-plot like view in Postscript files, and an interactive viewer is in development.