Tuesday, 28 July 2015

All hail the God of PCR

At a particularly low moment in my PCR optimisation process, I created a shrine on top of a cupboard. It was dedicated to a god of PCR. I'm against the non-critical thinking that leads to religion and faith, so you can probably imagine how low I had to be to create a shrine. But I thought that if, despite all evidence and logic, there was a god of PCR...I had better start worshipping it.

Since then, everything has pretty much gone to plan. I have had more samples work than not work, and I've begun to assemble a pretty comprehensive tree of Cantuaria. It has mostly shown me that the current taxonomy of Cantuaria places them into different taxonomic groups (species, genus) than their genetics do. But there is still a bit of work to do with BEAST, the program I am using the build the trees.

I've decided that optimising the stubborn samples that I have left (which all seem to require completely different PCR conditions) would be a waste of time and money. I have enough for my tree, and it would be nice to get them all to work, but I was meant to have completed this six months ago. I've got three new genes to work with as well (thanks to a friend) which I'm going to use to create a different phylogeny with fewer samples, but more sequences per sample (an idea which I picked up from the Evolution conference). That is not going to involve optimisation, though. If stuff doesn't work, screw it, I'm out of there. I have more ecological fish to fry.

So, all that I have left to do in the lab is...
-extract DNA from the 43 remaining spiders that I have collected
-PCR for all three genes these 43 spiders
-CO1 PCR a few more samples
-PCR the new 3 genes (I'm aiming for 14 sequences - two from each major clade).

If the PCRs for the 43 remaining spiders don't work, I'll try adding BSA (a magic liquid extracted by squeezing cows), doing a gradient PCR, and using MyTaq, which was recommended to me by another person who studies the same family of spiders (but seems a little sensitive for most of my samples). But that is the limit to which I will go to make them work.

There is a light at the end of the tunnel. Wish me luck!

No comments:

Post a Comment