Phylogenetic Tree With Dna, Binary And Multi Data - Raxml Update

Question

[Pavid]

Hey!

I've made a question about supertrees.
My goal is to integrate data from SNPs (DNA), MIRUs (multi -range around 1 to 25) and Spoligotype (binary) and eventually RFLPs, into a single tree.

@Dan Gaston suggested to use RAxML to obtain the same output.

I've been reading and for I what I understood I need the give a file with all the data (example file that I've created to test) and a partition file.
(I'm testing only with binary and dna data, for now)

In the command line:

./raxmlHPC-PTHREADS-SSE3 -m BINGAMMA -q partition_file.txt -s phylip_file.phy -n Test -T 2

but it gives me a error:

ERROR: Bad base (2) at site 47 of sequence 1
Problem reading alignment file

UPDATE

phylip file with the three types
partition file

command line and error:

./raxmlHPC-PTHREADS-SSE3 -m MULTIGAMMA -q test_part_all.txt -s test_all_types.phy -K ORDERED -n Multi -T 2

Multi State Error, characters must be used in the order they are available, i.e.
0, 1, 2, 3, 4, 5, 6, 7, 8, 9, A, B, C, D, E, F, G, H, I, J, K, L, M, N, O, P, Q, R, S, T, U, V   
You are using the following caharcters: 
0 1 2 3 4 A C G T

This is probably something easy that I'm not seeing.

Thank you guys!

[Pavid]

well I think so. IN my previous question I spoke about doing a supertree cause I though it was the best solution. I'm working my genotypes from M. tuberculosis. And we have SNP, MIRU, RFLP and spoligotype. So I wanted to integrate these methods into a single tree, as @Gaston suggested I've tried RAxML but I'm doing something wrong

[2184687-1231-83-]

just as a clarification, is consensifying different trees obtained with different methods into one final tree an option for you?

[Kumar]

Dear Pavid, Could you suggest me any tool for generating partition file. I have a alignment file which constitute 2000 genes and need to find the position of each gene from the alignment file. Is it possible to do by any tool?

Answer 1

[DG]

Your first partition looks like it is multi-state data, not binary. Binary data can only be 0, 1, and ? (maybe - I'm not sure). Anyway you have other state characters in there which is what it is complaining about, the first incorrect character for binary data is the character 2 at position 47 of the first sequence.

[Pavid]

yes, you're right. Its multi data. but for multi data this command should work no? ./raxmlHPC-PTHREADS-SSE3 -m MULTIGAMMA -q test_part.txt -s test_bin_dna.phy -K MK -n Test -T 2 it gives me the same error

[Pavid]

ok, I've corrected the positions and integrate them all in one file. now it gives me this error: Multi State Error, characters must be used in the order they are available, i.e. 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, A, B, C, D, E, F, G, H, I, J, K, L, M, N, O, P, Q, R, S, T, U, V You are using the following caharcters: 0 1 2 3 4 A C G T

[Pavid]

do I have to make a script to transform the data?

[DG]

Your partitions are marked wrong in the partitions file. Your Multistate data doesn't go all the way to 99. Double check your counts and how you have it set up.