发信人: happyzhb (~qualify过了，hia hia~), 信区: Biology
标 题: Re: ClustalW help-NEED EXPERTS!
发信站: Unknown Space - 未名空间 (Tue Oct 14 09:22:45 2003) WWW-POST

【 在 feizj (cornell) 的大作中提到: 】
: I have more than 2,000 groups need to be aligned, that's why I can't use
: public web interface (although they have many options but I can get all more
: options using command line) but to write shell scripts to run my alignments
in
: batch. I can't find the options to set score for mismatch and match
nucleotide
: in several multiple sequence alignment programs such as ClustalW, T-coffee.
: The default values (1.9 for match and 0 for mismatch in ClustalW) do not fit
: my case.
:
: Desperately need some experts to tell me how to do this? Or let me know any
: multiple sequence alignment program which has options for setting gap open
: penalty, gap extension penalty, score for mismatch and match nucleotide.

do you mean sth like this?
clustalw input.aln -gapopen=8.0

what software are you using? PC DOS stand alone clustalw or GCG? Unix or
windows? I am no expert, but I think this not that difficult, if you have an
ansi c compiler, I am sure you can do this easily.

For example, all parameters(like listed partially below):
*********************
/*
multiple alignment parameters
*/
float dna_gap_open = 15.0, dna_gap_extend = 6.66;
float prot_gap_open = 10.0, prot_gap_extend = 0.2;
sint profile_type = PROFILE;
sint gap_dist = 4;
sint output_order = ALIGNED;
sint divergence_cutoff = 30;
sint matnum = 3;
char mtrxname[FILENAMELEN+1] = "gonnet";
sint dnamatnum = 1;
char dnamtrxname[FILENAMELEN+1] = "iub";
char hyd_residues[] = "GPSNDQEKR";
Boolean no_weights = FALSE;
Boolean neg_matrix = FALSE;
Boolean no_hyd_penalties = FALSE;
Boolean no_var_penalties = TRUE;
Boolean no_pref_penalties = FALSE;
Boolean use_endgaps = FALSE;
Boolean endgappenalties = FALSE;
Boolean reset_alignments_new = FALSE; /* DES */
Boolean reset_alignments_all = FALSE; /* DES */
sint output_struct_penalties = 0;
sint struct_penalties1 = NONE;
sint struct_penalties2 = NONE;
Boolean use_ss1 = TRUE;
Boolean use_ss2 = TRUE;
sint helix_penalty = 4;
sint strand_penalty = 4;
sint loop_penalty = 1;
sint helix_end_minus = 3;
sint helix_end_plus = 0;
sint strand_end_minus = 1;
sint strand_end_plus = 1;
sint helix_end_penalty = 2;
sint strand_end_penalty = 2;
Boolean use_ambiguities = FALSE;
*********************************************
should be stored in a file called param.h, all you need to do is to change
them according to your needs and remember, after you change those, RECOMPILE
the program.

Good luck to your research!

:
:
: 【 在 bigfatliar (Chico Verde) 的大作中提到: 】
: : Can't you just go to this website and do all your analysis there? There
are
: a
: : bounch of options.
: : http://www.ddbj.nig.ac.jp/E-mail/clustalw-e.html
: :
: : 【 在 worryless (无忧) 的大作中提到: 】
: : : 太专业了，得问major in bioinformatics的人
: : : 【 在 feizj (cornell) 的大作中提到: 】
: : : : Does anyone know how to set the score for mismatch (negative value)
and
: : : match
: : : : (positive value) nucleotides, respectively, in clustalW (command
line)?
: : : :
: : : : Does anyone know how to make a DNA matrix file for clustalW?
: : : :
: : : :
: : : :
: : :
: : :
: :
: :
:
:


--
喂！盘丝洞不要乱闯
※ 修改:·happyzhb 於 Oct 14 09:22:45 修改本文·[FROM: 130.219.]
※ 来源:．Unknown Space - 未名空间 mitbbs.com．[FROM: 130.219.]