I am working on a data set for class and I was told that the sequences were generated using the Titanium chemistry. When I try and use the order=B I get no data in my trim.flows file and everything is sent to the scrap.flows file. When I run order=A (default) then I get data in my trim.flows. Additionally, mothur prompts me to use the lookup_Titanium.pat file instead of the GSFLX.
My question is, if I cannot use order=B (for titanium chemistry) why does mothur require me to use the titanium lookup (which is for the titanium chemistry)?
Both order=A and order=B are for Titanium. Order A is the old way in which the FLX run has always the flow order TACGTACG…, while in B the order of the nucleotides is random which should improve read length. Mothur needs to know which of the two schemes was used so that it can properly identify the results.
Looks like in your run the classical flow order A was used with your FLX Titanium run so nothing to worry about.
All seems to be going well now and it is nice to know these details for future studies. But what order would be used if the data was generated prior to the release of the titanium chemistry?
Why then are there several versions of a lookup file? If titanium can be used for all current and past chemistry’s, why have the others as an option to use?
Originally it was always a cyclic flow pattern with the 454 system, which would be order=A. In 2012 there was a software upgrade that gave the option to use the acyclic flow pattern (order=B); I don´t think that this included any hardware changes and it could be used with FLX+ or older versions of the machine like Titanium (not hundred percent sure here, we were already upgraded to FLX+ when the update came). So, before the second half of 2012 it would always have been order=A and after that it might have been either A or B depending on the chosen setting.
Why have different lookup files? You’d have to ask Chris Quince - we stole his and posted them on our wiki in a format that will work with mothur. It was only after we started playing with GS FLX+ data that we realized it didn’t make a big difference.