cat specific columns in 3 files - awk

Question

I have 3 files such as :
file1_file:
scaffold_159 625 YP_009345712 0.284 447 289 9 96675 95377 196 625 6.963E-38 158
scaffold_159 625 YP_009345714 0.284 447 289 9 96675 95377 196 625 6.963E-38 158
IDBA_scaffold_24562 625 YP_009345713 0.464 56 20 2 2549 2686 10 65 7.513E-03 37
file2_file:
scaffold_159 625 YP_009345717 0.284 447 289 9 96675 95377 196 625 6.963E-38 158
scaffold_159 625 YP_009345718 0.284 447 289 9 96675 95377 196 625 6.963E-38 158
IDBA_scaffold_24562 625 YP_009345719 0.464 56 20 2 2549 2686 10 65 7.513E-03 37
file3_file:
scaffold_159 625 YP_009345711 0.284 447 289 9 96675 95377 196 625 6.963E-38 158
scaffold_159 625 YP_009345723 0.284 447 289 9 96675 95377 196 625 6.963E-38 158
IDBA_scaffold_24562 625 YP_009345721 0.464 56 20 2 2549 2686 10 65 7.513E-03 37
and I would like to only get the 3th part of the 3 file in a single new_file.txt.
Here I should get :
YP_009345712
YP_009345714
YP_009345713
YP_009345717
YP_009345718
YP_009345711
YP_009345723
YP_009345721
From now I tried:
cat file_names.txt | while read line; do cat /path1/${line}/path2/${line}_file > new_file.txt; done
in file_names.txt I have :
file1
file2
file3
but I do not know how to extract only the 3th column...
Ps: the 3 files are not in the same directory :
/path1/file1/path2/file1_file
/path1/file2/path2/file2_file
/path1/file3/path2/file3_file

Answer 1

EDIT: After chatting with OP came to know that his/her files could be on different locations, so in that case could you please try following. Considering that you have an Input_file where all file names are there. I am yet to test it.
file_name=$(awk '{val=(val?val OFS:"")"/path1/" $0 "/path2/" $0} END{print val}' file_names.txt)
awk '{print $3}' "$file_name"
OR
awk '{print $3}' $(awk '{val=(val?val OFS:"")"/path1/" $0 "/path2/" $0} END{print val}' file_names.txt)
You could use awk here.
awk '{print $3}' /complete_path/file1 /complete_path/file2 /complete_path/file3

Answer 2

I think it can be simpler with
$ sed 's_.*_"path1/&/path2/&"_' filenames | xargs awk '{print $3}'
awk will be called only once.

Answer 3

So you have a file fnames.txt with hundreds of strings:
str1
str2
str3
str4
...
and each string represents a file located in
/path1/${str}/path2/${str}_file
where ${str} is a value from file fnames.txt.
Now you want to read the third column, from the third file only:
$ str="$(awk '(NR==3){print; exit}' fnames.txt)"
$ file="/path1/${str}/path2/${str}_file"
$ awk '{print $3}' "$file" > new_file.txt
Always remember the KISS principle