trimming bases

[hv4]

Hi,

When trimming reads I found Perl to be really slow for large files.

I wrote this C program to trim at the left and right ends.

Does this look OK to you?

#include<stdio.h>

void main(int argc, char *argv[])
{
FILE *fp_in, *fp_out;
char buffer[1000];
int count_lines;
int left_cut;
int right_cut;

if (argc!=4)
{
printf("Need input and output files and cuts, here's the order:\n");
printf("Input, output, left cut, right cut (reference from the left)\n");
return;
}

fp_in = fopen(argv[1], "r");
fp_out = fopen(argv[2], "w");

sscanf(argv[3],"%d",&left_cut);
sscanf(argv[4],"%d",&right_cut);

//printf("\n%d, %d\n", left_cut, right_cut);

count_lines = 0;

while(!feof(fp_in))
{
fgets(buffer, 1000, fp_in);

if ((count_lines%2)==0)
{
fputs(buffer,fp_out);
}
else
{
buffer[right_cut]=0;
fputs(&buffer[left_cut],fp_out);
fputc('\n',fp_out);
}

count_lines++;
}


fclose(fp_in);
fflush(fp_out);
fclose(fp_out);

}

[bioBob]

Don't forget to remove the mate if your trimming removes the whole read and you are looking at paired end data.

[westerman]

1) Assumes the arguments are what they are (e.g., file names)
2) Assumes reads are less than 1000 bases
3) Assumes reads are strictly header line followed by a single line of bases.
4) Assumes reads are all the same length
5) Assumes that the right cut position is not greater than the length of the reads.

I'd put, up-front, the assumptions so that people are not bitten by them. I'd also do more error checking.

[hv4]

Thank you for your feedback.

@bioBob:

The program does not remove any reads. Just trims them.

You can use it with paired reads since the program does exactly the same thing
for both reads.

@westerman

Your list is great. Indeed, the program is made to serve a narrow purpose.

To your list I would also add that the right_cut position specified as a parameter must be +1 than the one intended.

[bioinfosm]

how does this compare to cutadapt tool?