Hey,

For QC'ing GWAS sumstats with inconsistent RSIDs and missing INFO scores, plus liftover issues, I'd recommend the MungeSumstats R package from Bioconductor. It's robust, actively maintained (latest updates in 2024), and excels at standardizing sumstats across formats while handling these exact problems.

MungeSumstats can infer RSIDs from chromosomal positions and alleles using built-in references like SNPlocs, harmonize alleles, and perform liftover between builds (e.g., hg19 to hg38) via integration with rtracklayer. It also applies QC filters based on available columns, like allele frequency or p-values, even without INFO.

Install it like this:

if (!require("BiocManager", quietly = TRUE))
    install.packages("BiocManager")
BiocManager::install("MungeSumstats")

Then, run formatted <- MungeSumstats::munge_sumstats(path = "your_sumstats.txt", ref_genome = "GRCh37") – adjust parameters for RSID mapping and liftover as needed. Check the vignette for details: https://bioconductor.org/packages/release/bioc/vignettes/MungeSumstats/inst/doc/MungeSumstats.html

If you need a more automated pipeline, consider tidyGWAS (new in 2024), which separates validation from QC and supports harmonization.

This approach should standardize everything on RSIDs and sidestep liftover pitfalls better than GWASinspector.

Kevin