I have been thinking how to address this particular task but is way beyond my knowledge.
I have a reference sequence, something like this:

Code:

>Reference
AGAGAGACCTGGAGAGAGAGTGACGATGAGCAGTGACGATGACGTACGATAGCAGTAGACGCA

and a input.txt file with thousand of short sequences, something like this

Code:

>read1 ori 498
AGAGAGACCTGGAGAGAGAGT
>read2 ori-rep 500
AGAGAGACCTGGAGAGAGAGT
>read3 1-misma 456
GGAGAGACCTGGAGAGAGAGT
>read4 2-misma 456
TGAGAGACCTGGAGAGAGAGA
>read5 ori-rev 532
ACTCTCTCTCCAGGTCTCTCT
>read6 ori-rev-1-misma 499
ACTCTCTCTCCAGGTCTCTCC
>read7 medium 512
AGAGAGAGTGACGATGAGCAG
>read8 last 488
AGTGACGATGACGTACGATAGCAGTAGACGCA
>read9 last rep 488
AGTGACGATGACGTACGATAGCAGTAGACGCA
>read10 last gap 488
AGTGACGATGACGTACGATAGCAGTAGACGA
>read11 nomatch1 500
GGGGGGAAAAAGCGTGCGT
>read12 nomatch2 500
CCCCGGGATGACGATGACGATGACGATGACGATGAC
>read13 nomatch3 550
GGGGTGCGAAAAAACCCCCGGGGTGG
>read14 nomatch4 543
TTTTTTTTTTAAAAAGCCGCGCTTTTTTT
>read15 nomatch5 543
TTTTTTTTTTAAAAAGCCGCGCTTTTTAA

The output file should contain the following:
1. All sequences that match the reference sequence 100% (in my example, sequences 1, 2, 7, 8 and 9)
2. If a sequence does not match the reference, it should reversed and complemented (A=>T; T=>A; C=>G; G=>C), and run against the reference sequence for a second time. If it matches, it should be included in the output file as reversed/complemented sequence (sequences 5)
3. All sequences containing 1 or 2 mismatches should be included without changes (sequences 3 and 4)
4. All sequences that after being reversed and complemented contain 1 or 2 mismatches should also be included as reversed/complemented sequences (sequences 6)
5. All sequences missing 1 character (sequence 10)

Resulting in the following outfile

Code:

>read1 ori 498
AGAGAGACCTGGAGAGAGAGT
>read2 ori-rep 500
AGAGAGACCTGGAGAGAGAGT
>read3 1-misma 456
GGAGAGACCTGGAGAGAGAGT
>read4 2-misma 456
TGAGAGACCTGGAGAGAGAGA
>read5 ori-rev 532
AGAGAGACCTGGAGAGAGAGT
>read6 ori-rev-1-misma 499
GGAGAGACCTGGAGAGAGAGT
>read7 medium 512
AGAGAGAGTGACGATGAGCAG
>read8 last 488
AGTGACGATGACGTACGATAGCAGTAGACGCA
>read9 last rep 488
AGTGACGATGACGTACGATAGCAGTAGACGCA
>read10 last gap 488
AGTGACGATGACGTACGATAGCAGTAGACGA

The second outfile should be based on the first outfile. Here, I would like to assemble all sequences into one by overlapping the matching portions and name the new reference with the input file name. An "N" should be inserted if a variable position is found:

Code:

>input
NGAGAGACCTGGAGAGAGAGNGACGATGAGCAGTGACGATGACGTACGATAGCAGTAGACGCA

I know perl will probably be the best way to go. However, my understanding about perl is quite limited and I do not think AWK would be the best way to solve this task
Any help will be greatly appreciated