how to replace characters into digits
显示 更早的评论
i have to replace the each characters using the following digits s=ACGT
i have to replace as
'A' then 11
'C' then 00
'G' then 01
'T' then 10
1 个评论
Jan
2013-2-27
You write 11 without surrounding quotes. This isn't an accident, correctly? What do you want as output? How long is the input?
采纳的回答
Azzi Abdelmalek
2013-2-27
编辑：Azzi Abdelmalek
2013-2-27
clear
s='ACGT'
e=['11';'00';'01';'10']
in='AGCTAG' % Your initial data
out=in
for k=1:numel(s)
out=regexprep(out,s(k),e(k,:))
end
更多回答（3 个）
Jan
2013-2-28
And a lookup table:
seqIn = 'ACGTTGCA'
table = repmat('0', 2, 255);
table(1, 'AT') = '1';
table(2, 'AG') = '1';
result = reshape(table(:, seqIn), 1, []);
Does this work? I do not have access to Matlab currently.
3 个评论
Azzi Abdelmalek
2013-2-28
编辑：Azzi Abdelmalek
2013-2-28
Yes It works
Azzi (regexprep),Azzi (strrep),Cedric,Jan execution time in this order
125000 character
Elapsed time is 14.543628 seconds.
Elapsed time is 0.007702 seconds.
Elapsed time is 0.005915 seconds.
Elapsed time is 0.006354 seconds.
Actually STRREP still wins up to me. The solution based on ISMEMBER is profiled between REGEXP and the two solutions based on array indexing. But the latter require much more memory during the indexing operation than STRREP. You start seeing that with ~1e8 chars. On my laptop with 8GB RAM for example, only STRREP can treat more than 2e8 chars without swapping (or being killed by Process Lasso).
Jan
2013-2-28
@Azzi: Is this a typo?! Your function needs 14 secs with REGEXPREP and 0.007 secs with STRREP? Then my minor suggestion caused a speedup of a factor 1900? Wow, this would be the most efficient suggestion I ever gave. And it would be a strong hint to warn for the low efficiency of regexprep in this forum.
If you need to process long sequences, you might want to optimize a little the efficiency.. a MEX-based solution would be most efficient I guess, but here is one way you could go using basic MATLAB only..
If you want to replace character 'A' with a numeric array (1,1) and so on, you can do:
aa = 'ACGT' ;
seq = 'AAGCTCAGGTTCA' ;
rep = zeros(2, max(aa), 'uint8') ;
rep(:,aa) = [1 0 0 1; 1 0 1 0] ;
result = reshape(rep(:,seq), 1, []) ;
This outputs the numeric array:
result =
1 1 1 1 0 1 0 0 1 0 0 0 1 1 0 1 0 1 1 0 1 0 0 0 1 1
If you want to replace character 'A' with characters '11' and so on, you can do:
aa = 'ACGT' ;
seq = 'AAGCTCAGGTTCA' ;
rep = zeros(2, max(aa), 'uint8') ;
rep(:,aa) = ['11'; '00'; '01'; '10'].' ;
result = char(reshape(rep(:,seq), 1, [])) ;
This outputs the string '11110100100011010110100011'.
EDIT: note that there are slightly different ways of doing this a little slower but with a more memory-friendly approach.
Cheers,
Cedric
Jos (10584)
2013-2-28
Here is a simpler approach than looping over REGEXPREP or STRREP:
seqIn = 'ACGTTGCA' % input sequence
letters = 'ACGT' ;
symb = {'11','00','10','01'} ; % stored as a cell array of strings!
[tf,idx] = ismember(seqIn,letters) ;
seqOut = [symb{idx}]
类别
在 帮助中心 和 File Exchange 中查找有关 Characters and Strings 的更多信息
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!
Translated by 
