Messing around with the mitochondria
OK, just a bit more procrastination. I wanted to try one more time to look at mitochondria SNPs. Mapping to the PA42 mitochondria was seeming a bit messy, so I wanted to try a different approach. I looked for D84A scaffolds that had decent blast lengths to the PA42 mitochondrion (over 1kb). I then took the sequence of those D84A scaffolds and blasted them to the nucleotide database using blastn on NCBI's website. I found three D84A scaffolds that had blast hit lengths to PA42 of over 1kb, and when blasted on NCBI had really good matches to Daphnia mitochondria. These three scaffolds vary in length from 2407-3314, and sum to 8246bp, which is a little over 50% of the size of the mitochondria in PA42. I then extracted the SNPs from just those contigs using SNPrelate and did a PCA on all the 20162017 data (without D Barb). After filtering for maf=0.15 and missing=0.5, I had 15 SNPs. I did not do LD pruning. Here is the result of the PCA, with the size of the circle matching the frequency at which that mitochondria was found.
What is up with this??? Why are there multiple haplotypes in the 2016 data, but in all the 2017 ponds there is only a single haplotype??? Is it possible that this is due to some sort of microbial contamination? 2016 data was not treated with antibiotics, while 2017 was... Not sure if this is a possibility or not, or how to test it. I did look at how this overlaps with the nuclear data. For D8 in 2016, the individual that has the same haplotype as one of the 2017 D8 super clones does have the same mitochondrial haplotype as that seen in all the 2017 D8 clones. There was also one small super clone in D8 2016, with four individuals, that is very similar to one of the super clones D8 2017, and all four individuals also have that same mitochondrial haplotype. The unique clones in D8 2016 have a mix of the 2017 haplotype (n=8), or unique mitochondrial haplotypes (n=6, or a shared otherwise unique haplotype (n=2))
If the PCA is true, one thing this suggests is that looking at cytonuclear linkage is not going to be possible, since there is only one mitochondrial haplotype in the 2017 data. But why is there only one mitochondrial haplotype?
What is up with this??? Why are there multiple haplotypes in the 2016 data, but in all the 2017 ponds there is only a single haplotype??? Is it possible that this is due to some sort of microbial contamination? 2016 data was not treated with antibiotics, while 2017 was... Not sure if this is a possibility or not, or how to test it. I did look at how this overlaps with the nuclear data. For D8 in 2016, the individual that has the same haplotype as one of the 2017 D8 super clones does have the same mitochondrial haplotype as that seen in all the 2017 D8 clones. There was also one small super clone in D8 2016, with four individuals, that is very similar to one of the super clones D8 2017, and all four individuals also have that same mitochondrial haplotype. The unique clones in D8 2016 have a mix of the 2017 haplotype (n=8), or unique mitochondrial haplotypes (n=6, or a shared otherwise unique haplotype (n=2))If the PCA is true, one thing this suggests is that looking at cytonuclear linkage is not going to be possible, since there is only one mitochondrial haplotype in the 2017 data. But why is there only one mitochondrial haplotype?
Hmm. that is super interesting. It would probably be good to look at mitochondrial haplotypes given the limited number of SNPs. It might give more information about whether the 2016 results are a artifact of bacterial contamination.
ReplyDeleteBut, even if we ignore the little dots in D8 and D10, there does seem to be two distinct mitochondrial lineages between those two ponds, right? And the W ponds are also distinct from the D ponds?
Maybe we can talk about looking at the mitochondrial haplotypes on Friday.
ReplyDeleteYes, there are distinct mitochondrial lineages between D10 and the D8 and surrounding ponds. And the W ponds are distinct from one another and the D ponds.