Error correction about atlas HOT 4 CLOSED

Error correction is optional and if errors are present in sequences it will affect their ability to map to decontaminant reference sequences. That was my thought anyways. SPAdes will perform an additional step of error correction in its protocol.

from atlas.

For publication, you normally want to upload raw reads. But at least for human metagenomics one should remove the reads from the host for confidentiality reasons. That was a second thought to put the decontamination step first and all the quality filtering and error correction later?

from atlas.

That definitely makes sense in that context.

We now force users to either start with single- or paired-end after de-interleaving their own reads using reformat.sh. It seems like a simple switch and I'm open to it.

from atlas.

The rule order has been updated to reflect corrections highlighted in the original issue.

from atlas.