3.8 years ago by
BWA is a software package for mapping reads against a large reference genome. BWA is fast but not as sensitive as other programs, ex., blast or needle. To achieve great speed, BWA requires the reads to show low-divergence from the reference (see program description at http://bio-bwa.sourceforge.net/) and uses some heuristics (tricks) that assume that there is a long stretch of DNA that matches exactly between the ref and the reads. This is called a seed. Depending on the version of BWA you are using, this can be between 19 to 32 bases by default. In your case, you are most likely not getting any hits before your references are shorter than the seed length.
If you are looking for exact matches, then you could use grep, as suggested by @h.mon. This would require putting the sequence and the id on the same line, so that you can retrieve the hits.