3.8 years ago by
You can convert a fasta file to BAM format, but it will not have chromosome mapping information.
In order to obtain that, map the full dataset then filter the results for hits to the target chromosome. Then go back and extract just the fasta sequences for those hits, creating the final .fastq dataset to use in your analysis as the input. This creates slightly skewed input - only sequences that map will be retained (meaning, unmapped sequences will not be a part of the input, as some fraction would normally be). This may or may not matter to you, and you could always seed back in some unmapped sequences at the same fraction found in the original dataset.
Alternatively, you could create a custom reference genome with just the target chromosome and use that when you map. The job will execute quicker. However, you will almost certainly get slightly different results. Perhaps try both and see which works best for you on one dataset, then use that method with the others.
Best, Jen, Galaxy team