Question

I would like to having your helps on keeping only the first letter for any strings in fields of a tab-delimited file.

I think sed can do this job. I tried, but I failed. I expect to having your advice and directions.

Thanks in advance.

A dummy example:

my_file, (tab-delimited)

1   11656   TCAG    .   TCAG    TCAG    .   TCAG    TCAG    TCAG
1   11660   CT  .   CT  CT  .   CT  CT  CT
1   11662   ATGG    .   ATGG    ATGG    .   ATGG    ATGG    ATGG
1   11680   A   .   A   A   .   A   A   A
1   11732   C   .   C   C   .   C   C   T
1   11742   T   .   T   C   .   T   T   T

What I want:

1   11656   T   .   T   T   .   T   T   T
1   11660   C   .   C   C   .   C   C   C
1   11662   A   .   A   A   .   A   A   A
1   11680   A   .   A   A   .   A   A   A
1   11732   C   .   C   C   .   C   C   T
1   11742   T   .   T   C   .   T   T   T

Code I tested:

# (1) workable for only one string
echo abcd123 | sed 's/\([a-z]\).*/\1/'
# (2) not work for my data file
sed 's/\([a-z]\).*/\1/' my_file
sed 's/\([a-z]\).*/\1/g' my_file

Answer 1

Manny’s solution but a bit more generic

sed 's/\([A-Z]\)[A-Z]*/\1/g' my_file