I’m building a snakemake pipeline to align some NGS data, and some of the samples are single end and others are paired end, which require slightly different analyses. My solution was to declare a sample_sp wildcard to use for rules that are specific to a certain type of data. I want this wildcard to not end with _pe, as that is how I am marking files to be used in paired-end experiments.

I saw these answers give solutions using negative lookahead or lookbehind regexes, but snakemake doesn’t seem to match the wildcard using these regexes to any of my data files.

Is there any way to get snakemake to use negative lookahead/lookbehind? Or is there another way I can define a wildcard constraint to not end with a certain string?