Help in Python and AWK: Printing absolute difference value between consecutive rows values, if column1 value is same

Question

[tinkuhim007]

Need help in Python and AWK: Printing least difference value between consecutive rows values, if column1 value is same Being a biologist, not able to do it. Your help is highly appreciated.

Input File: <tab delimited file>

Chr1     A     10
Chr1     B     13
Chr1     C     12
Chr2     D     12
Chr2     E     14
Chr2     F     11

Description: In column 1, chr1 == chr1; so A-B, A-C, and B-C. Further, Chr2==Chr2 so same iteration will be followed.

Output File 1: <tab delimited file>
A     B     3
A     C     2
B     C     1
D     E     2
D     F     1
E     F     3

[Ram]

Your output assumes each member of column 2 is unique to the chromosome. I'm assuming column 2 has genes and column 3 has some sort of coordinates and you're looking to list intergenic distances of some sort?

[h.mon]

Be careful what you wish for, because you may get it: the number of differences you want is given by the formula

n! / ( k! * ( n - k )! )

Where nis the number of elements ( A, B, C, ... ) per chromosome, and k = 2. For 10 elements on a chromosome, you get 45 differences, for 100 elements per chromosome, you will get 4950 differences, and so on. My cell phone calculator cannot display the results for 1000 elements, as the number is already too big for it.

[Ram]

It's 499500 combinations for 1000 elements [C(n,2)]. I think we have ourselves an XY problem here.

Answer 1

[Kevin Blighe]

Hello,

Here is an awk command that will do it for you per chromosome, tinkuhim007:

cat test
Chr1    A   10
Chr1    B   13
Chr1    C   12
Chr2    D   12
Chr2    E   14
Chr2    F   11

awk '{arr[$1","$2]=$3} END { \
    for (char1 in arr) { \
        for (char2 in arr) { \
            split(char1, charArr1, ",") ;
            split(char2, charArr2, ",");
            if ((char1 != char2) && (charArr1[1] == charArr2[1])) { \
              print charArr1[1]"\t"charArr1[2]"\t"charArr2[2]"\t"arr[char2]-arr[char1]}}}}' test

Chr1    A   B   3
Chr1    A   C   2
Chr1    B   A   -3
Chr1    B   C   -1
Chr1    C   A   -2
Chr1    C   B   1
Chr2    D   E   2
Chr2    D   F   -1
Chr2    E   D   -2
Chr2    E   F   -3
Chr2    F   D   1
Chr2    F   E   3

You didn't indicate a rule for order of subtraction. If you don't want negative values anywhere in output, you can just add an extra if statement:

awk '{arr[$1","$2]=$3} END { \
    for (char1 in arr) { \
        for (char2 in arr) { \
            split(char1, charArr1, ",") ;
            split(char2, charArr2, ",");
            if ((char1 != char2) && (charArr1[1] == charArr2[1])) { \
              result = arr[char2]-arr[char1] ;
              if (result > 0) { \
                  print charArr1[1]"\t"charArr1[2]"\t"charArr2[2]"\t"result}}}}}' test

Chr1    A   B   3
Chr1    A   C   2
Chr1    C   B   1
Chr2    D   E   2
Chr2    F   D   1
Chr2    F   E   3

If you want to understand how this is working, then let me know.

Kevin

[tinkuhim007]

Dear Kevin
Thanks a lot for helping me. I very new to the awk. It will be great if you could explain me line wise the code. And one more thing is there if you could kindly add in this code, only least distance value appears between two points in the second output file.

Output file <Tab delimited>
Chr1    A     C     2
Chr1    B     C     1
Chr1    C     B     1
Chr2    D     E     2
Chr2    E     D     2
Chr2    F     E     3

Thanks a lot.