3 c14@sddlZddlZddlZddlZddlZddlZdZdZedkrHe dddd Z d d Z d d Z ddZ ddZddZddZddZdddZdS)NzGraham Alvare, Brian FristenskyzSun Mar 18 14:57:59 CDT 2012Z__main__zPHYLIP Library moduleTFcCst|d}tdd}|j|j|jtdd}|j|j|j|rtdd}|j|j|j|r|jd|jd|jd|jd|jd|jd|jd|jd S)a Merge messages from MSGFILE to the outfile generated by a Phylip program. If extra is included as a parameter, this appends a message indicating that the output is a consensus tree, whose branch lengths are the bootstrap values, NOT the actual branch lengths. wMSGFILEroutfilezoutfile.consense z<>>>> THIS TREEFILE IS A CONSENSUS TREE, WHOSE BRANCH LENGTHSz5>>>> ARE BOOTSTRAP VALUES, NOT ACTUAL BRANCH LENGTHS.z>>>> TO GENERATE BRANCH LENGTHSz6>>>> USE TREE FILE AS INPUT FOR DNAML OR OTHER PROGRAMz>>>> USING THE USERTREE OPTIONNrrz=>>>> THIS TREEFILE IS A CONSENSUS TREE, WHOSE BRANCH LENGTHS z6>>>> ARE BOOTSTRAP VALUES, NOT ACTUAL BRANCH LENGTHS. z >>>> TO GENERATE BRANCH LENGTHS z7>>>> USE TREE FILE AS INPUT FOR DNAML OR OTHER PROGRAM z>>>> USING THE USERTREE OPTION )open writelines readlinesclosewrite)Z outfile_pathZconsenseZextraZ h_OUTFILEZ h_MSGFILEZh_smalloutfileZh_outfile_consenser '/home/psgendb/BIRCHDEV/script/phylip.py merge_msgs(           rcCs&t|d}|jjd}|j|S)zIRead first line of a Phylip file to find out how many sequences there arerr)rreadlinesplitr )infileZh_infileNUMSEQr r r get_numseq:s rcCsPt|}|dkr2||kr2|jd|j|dnd}|jdt|d|S)z-Make sure OUTGROUP is not greater than NUMSEQorz OUTGROUP = zo )intr str)OUTGROUPrcomfile h_msgfileZ tempoutgroupr r r do_outgroupBs rcCs$tjdd}t|ddd}|S)a1Generate a random integer as needed by Phylip programs. These numbers are used by Phylip programs as seeds for a random number stream. They must be odd, in the form 4n + 1. Return value is an integer between 0 and 2e16 -1 (which is 65535). Although the Phylip Main document claims that 32-bit random numbers are acceptible, at least one program, PARS, will only take 16-bit random numbers. Use floor division by 4 to generate a whole number quotient, multiply the quotient by 4 to generate an even number, and add 1 to make it odd. rir)randomZrandintr)ZpseedZprandr r r phylip_randomNs rc Cs>tj} dttj} tj| tj|tjj| dtj | t dd} t |dksdt |dkrhd}| j d| j t|dt |d kst |dkrd }t |d kr| j d | j t|d|d kr| j d nr|d kr| j d | j d nT|dkr| j d | j d n4|dkrD| j d | j d | j d n |dkrN|dkrn|dkrh| j dn|dkr| j d|dkrh| j dn|dkr| j d| j dd}n|dkr| j d| j d| j dd}n~|dkr"| j d| j d| j d| j dd}nF|dkrd| j d| j d| j d| j d| j dd}nd}| j d| j t|d|dkr| j d| j dt } | j t| d| jt dd } tjdg| d}|j| j|dkrtjd tjj| | ntjd!tjj| | tj | tj| d"d#S)$aRun Seqboot. Notes: SEQBOOT reads interleaved sequences by default, but can read sequential files using the "I" setting. By default, SEQBOOT writes datasets to outfile, but will write weights to 'outweights' if you set the "S" option. Weight files are always sequential. zSEQBOOT.rZSeqbootComfilerdz% rrzB mzd rRgsbyeszs dzj psnopopwZrewzr zi zy seqboot)stdinZ outweightsrTN)osgetcwdrgetpidmkdirshutilcopypathjoinchdirrrr rr subprocessPopenwaitmovermtree)INFILEDATATYPERSEEDMETHOD REPLICATESPERCENT BLOCKSIZEZ OUTWEIGHTS OUTFORMATOUTFILESTARTDIRTEMPDIR comfile_hZtempseedpr r rr-_s                                                r-cCsVtt}|jdt|d|d|jd|j|d|jt|ddS)zJumble - When multiple datasets are analyzed, protpars automatically jumbles, and prompts for a random number seed for jumbling. Otherwise, jumbling must be explicitly set.zJUMBLING SEQUENCE ORDER z ITERATIONS, SEED=rjNzj )rrr )rrNUMJUMZ tempjseedr r rjumbles   rLc Csht}t|}t|} |dkr,|jdn8|dkr|dkrv|jd| d|dt|dkr|jd |dn|jd | d|dt|d kr|jd |d dtjddtd||||||d|d n|dkrd|dkr |jd| d|dn@|dkr0|jd| d|dn|jd| d|dtd||||||d|d dS)zRun SEQBOOT to generate resampled datasets. Output is weights, used by Phylip programs to generate datasets on the fly. If you want SEQBOOT to generate actual datasets instead of weights, use weightless_resample instead.n rr&r(zRESAMPLING: Bootstrap, z REPLICATES, SEED=rzResampling in blocks of z$RESAMPLING: Delete-half Jacknifing, r!zPartial Resampling: zpercent of sites sampledz infile.temprr'Zweightsr)r+r,z0RESAMPLING: Permute species for each character, z%RESAMPLING: Permute character order, z$RESAMPLING: Permute within species, r*Nz )r&r()r)r+r,)rrr rr3Zcopyfiler-) r@rBrArCrr>rDBseedBseedStrRepStrr r r stdresamples,      rRc Cs t}t|}t|} |dkr*|jdn|dkr|dkrX|jd | d |dn|dkr||jd | d |dn|dkr|jd | d |dnb|dkr|jd | d |dn|jd| d |dt|dkr|jd|ddtd||||||d|d dS)z Generate terminal input for running seqboot to generate randomized sequence files, rather than weight files and run seqbootrMrNrr&r(r)r+r,z0RESAMPLING: Permute species for each character, z REPLICATES, SEED=z%RESAMPLING: Permute character order, z$RESAMPLING: Permute within species, zRESAMPLING: Bootstrap, z$RESAMPLING: Delete-half Jacknifing, r!zPartial Resampling: zpercent of sites sampledz infile.tempr*rNz )r&r(r)r+r,)rrr rr-) r@rBrArC msgfile_hr>rDrOrPrQr r rweightless_resamples$  rTcCs|rttjj|dd}ntj}t|d}|j}|jd}|djdd krx"|D]}|jdd krV|d}qVW|j t |dx|D]}|j |qW|jd S) zUsed by dnaml.py and protml.py. Make sure that treefile begins with number of trees on first line of file. If first line in file has parentheses, the number must be added.Zintreearr(r;rNrX) rr/r5r6sysstdoutr r findr r)ZUFNrGrZinfile_hZufn_hZ ufn_linesZsc_countliner r rufn4s    r])TF)T)r/os.pathrr3r8rY __author__Z__date____name__printrrrrr-rLrRrTr]r r r rs& %  (