grouping exons together per gene?

[Sonia NOSRATINIA]

Hello. I'm new to partitionfinder and was not able to find information on this in the manual. When creating a data block for a large set of data with multiple little exons per protein coding gene, when attempting to separate out the codon positions, is there a way to put all of the exons for the same gene in one group and all of the introns in another group or would I need to have a block for each individual exon with codon positions specified and a block for each individual intron with no codon position specified? In other words, I'm trying to learn what the syntax is for defining one data block for each codon position in each gene that has many exons scattered through the gene (if such a syntax exists). I should mention that I am working with a large concatenated alignment in which ambiguous regions of each locus have been deleted but these deletions were only in the intron regions. The exons are intact although in the cases where the exon is in the beginning of the alignment or the end of the alignment we may not have the complete exon. Thanks!

[Rob Lanfear]

Hi Sonia,

Yes, you can group non-contiguous sites together in the data blocks, and the approach you suggest (all the exons of one gene grouped) is a sensible one. 

Here's an example of how to do it (where positions 101-205 might be an intron):

gene1_pos1 = 1-100/3 205-400/3;

Hope that helps,

Rob

[Sonia NOSRATINIA]

Thank you, Rob. When not specifying the codon positions (say for example when grouping a series of introns or spacer regions), would the ranges be separated by just a space, for example 1-100 205-400?

Sonia