fastman: A fast algorithm for visualizing GWAS results using Manhattan and Q-Q plots

Visualization of GWAS summary statistics, specifically P-values, as Manhattan plots is widespread in GWAS publications, and many popular software tools are available, such as the R package qqman. But there is substantial need for further development, such as the handling of non-human data. We provide a new R package, fastman, with major additional capabilities. It handles genomes of non-model organisms, even those at a draft stage, i.e. contigs that havent been compiled to chromosomes. Non-numeric chromosome IDs are supported. It supports plotting of other genetic scores, such as FST, D statistics, selection statistics such as PBS, or other kinds of GWAS statistics such as beta. Importantly, negative or two-tailed values are supported in this package. We implement a heuristic algorithm that drastically reduces plotting time for huge datasets without any loss of visual precision, while allowing for many different data types and missing data. We provide substantial additional flexibility in highlighting and annotation. In summary, we have developed a package fastman in R for fast and efficient visualization of GWAS results and other genomewide scores using Manhattan and Q-Q plots. The package can create plots directly from association outputs by PLINK. Alternatively, it can produce plots from any R data frame with custom columns and is equipped to handle big datasets with fast plot generation. It is available for public use on https://github.com/kaustubhad/fastman.