Issues (WIP)

• SPAdes can fail during assembly of some kmer lengths in a manner that isn’t detected by HybPiper (looks like this is a bug in SPAdes), meaning that real contigs are missed for some loci.
• SPAdes can hang indefinitely at the kmer coverage model stage (see https://github.com/ablab/spades/issues/653), meaning the pipeline never finishes. Running SPAdes in single-cell mode (--SC) seems to fix this, but it can also create many more spurious contigs than running without single-cell mode, which can in turn sometimes cause errors in the loci sequence output.
• Detection of paralogs - it’s currently very rough, and will miss many paralogs.