The results of the STAR alignment for the trimmed reads were a little strange looking (especially in comparison to the stress-hardening samples, those look way better), so I tried aligning the raw reads and seeing if I got better results (higher alignment and less unmapped reads). Interestingly, it looks like trimming improved alignment rates (trimmed samples on left, raw reads on right):

In comparison, here is what the stress-hardening alignment looks like (for the Acer samples specifically):

In terms of what I mean by “better results”, I mean higher percentage alignment and overall higher read alignment in terms of number of bases. I’m not sure what a good number is for these but I think it is something like >50% alignment. Idk though.

Also, I’m not sure what a good alignment score is. For the stress-hardening samples, this is what the multiqc report looks like:

versus the Acer CCC samples:

I think overall I can move forward with samples that have greater than 30% alignment for the Acer CCC ones, and move on with all the stress-hardening samples and see how the gene counts come out. Then I need to try to align the unmapped reads to the symbiont genome (Symbiodinium fitti) to see if those get high alignment.

Also, I thought I wrote this somewhere already but I can’t find it, but it is important to note that the Acer CCC and the Acer stress-hardening were sequenced differently.

Acer CCC = 3’ Quantseq NovaSeq S2

Acer stress-hardening = NovaSeq S2 SR100

… Wait they were sequenced the same way lol. Nevermind. I guess that in and of itself is interesting. The CCC samples may have been worse quality because they were collected in the field? Or who knows. But that’s actually really interesting that they were pretty much sequenced the same way.