Calculating CG content between delimeters

Hi, I'm new to matlab and i'm having this problem. For exp:I have DNA sequences in one string separated by delimiter 'y'.

A= CCTGCGGAAGATCGGCACTAGAATAGCCAGAACCGTTTCTCTGAGGCTTCCGGCCTTCCCTCCCACTAATAATTCTGAGG y CCATCGGTAGCGCATCCTTAGTCCAATTAAGTCCCTATCCAGGCGCTCCGCCGAAGGTCTATATCCATTTGTCAGCAGACACGC y CCACCCTCGTGGTATGGCTAGGCATTCAGGAACCGGAGAACGCTTCAGACCAGCCCGGACTGGGAACCTGCGGGCAGTAGGTGGAAT

I need a loop that counts nucleotides(C,G) between delimeters 'y' separetly and puts answers to cell array.

b=0;
for i=1:length(A); 
   if A(i)~='y'; % if i is not equal delimeter
     b=b+1;   %  at first i'm just trying to all count nucleotides
              % i don't know how to put answer to cell 
   elseif A(i)=='y';
   % if i equals delimeter 'y' count nucleotides from new until next 'y' and put answer to cell 

   end
end
Any bright ideas ?

Answer 1

[Star Strider]

If you have the strsplit function, this is straightforward:

A = 'CCTGCGGAAGATCGGCACTAGAATAGCCAGAACCGTTTCTCTGAGGCTTCCGGCCTTCCCTCCCACTAATAATTCTGAGG y CCATCGGTAGCGCATCCTTAGTCCAATTAAGTCCCTATCCAGGCGCTCCGCCGAAGGTCTATATCCATTTGTCAGCAGACACGC y CCACCCTCGTGGTATGGCTAGGCATTCAGGAACCGGAGAACGCTTCAGACCAGCCCGGACTGGGAACCTGCGGGCAGTAGGTGGAAT';

Aseg = strsplit(A);
CGpos = strfind(Aseg(1:2:end), 'CG');

Without strsplit, it requires a loop:

yidx = [1 strfind(A, ' y ') length(A)];             % Find Delimiters
for k1 = 1:length(yidx)-1
    Aseg{k1} = A(yidx(k1):yidx(k1+1));
    if k1 > 1
        Aseg{k1} = Aseg{k1}(3:end);
    end
end
CGpos = strfind(Aseg,'CG');

Note that this finds the occurrences of 'CG' together. If you want individual occurrences of cytosine and guanine (and you’ve used strsplit), search for them individually:

Cpos = strfind(Aseg(1:2:end), 'C');
Gpos = strfind(Aseg(1:2:end), 'G');

To find the numbers of occurrences in each sequence (regardless of the way you calculated ‘CGpos’):

CGsum = cellfun(@length, CGpos);

So, use strsplit if you have it to avoid the loop. Note that to calculate CGpos, the argument to strfind is different if you use the loop or if you use strsplit.

[Rokas]

Thank you very much. And what do you sugest for adding Cpos and Gpos values into new array. CGsum = 'Cpos'+'Gsum'; something like this.

[Rokas]

CG=[Csum; Gsum]
CGsum=sum(CGsum);
oh my it was simple.

[Star Strider]

My pleasure!

(The most sincere experssion of thanks here on MATLAB Answers is to Accept the Answer that most closely solves your problem.)

—————

Note that for the rest of this, the variables assume:

% With ‘strsplit’:
A = {A};                    % ‘A’ As Cell Array Here
Qseg = strsplit(A{:});
Aseg = Qseg(1:2:end);

That said . . .

Don’t add ‘CGpos’, ‘Cpos’, or ‘Gpos’ to ‘CGsum’. They’re different variable types. The ‘CGpos’, (and ‘Cpos’, and ‘Gpos’) are cell arrays, and ‘CGsum’ (and ‘Csum’ and ‘Gsum’ respectively) are double arrays with the number of occurrences for each sequence in each element.

If you want to sum the ‘C’ and ‘G’ occurrences separately, I would keep the sums separate:

Cpos = strfind(Aseg, 'C');
Gpos = strfind(Aseg, 'G');

Csum = cellfun(@length, Cpos);
Gsum = cellfun(@length, Gpos);

then sum them individually if you want. In my code, ‘CGsum’, ‘Csum’, and ‘Gsum’ are entirely different. As written here, they give you the numbers for ‘CG’ (or ‘C’ and ‘G’ individually) in each sequence. If you want the total numbers in all sequences, just sum them:

SCsum = sum(Csum);
SGsum = sum(Gsum);

The same would be for the sum of the ‘CG’ sequences:

CGsum = cellfun(@length, CGpos);
SCGsum = sum(CGsum);

That should work without problems.