fasta sequence manupulation

Question

[poonam.bi01]

i have fasta file like this:

>1001796365 4.F.1.1.5
MDSIRPATFQIPAAVRELGWAALLLFFVLLSVHEWFSPPGWFGLLAILIFATQGALILTR
WPARQNFGWANRTTLLRSILVVSLVAWAPFLPAADSSALWIYGVACLIALILDGVDGKVA
>1002048002 2.A.4.2.8
MSPSRTARLYFLLVLDLLFFVLEISIGYAVGSLALVADSFHMLNDVVSLIIALYAIKLAA
SSTPTTRYSYGWHRAEILAALVNGVFLLALCFTITLEALERFFSTPEISNPKLIVLVGSL
>1002048004 2.A.4.5.2
IASDIRRILHRHGIHSSTIQPEYHPVRDTILEERSKDVNCLISCPPDSACCEVQACCPSY
AGT

header order in fasta sequence :

 >+first_id then+\t+second_id

i want my sequence in this formate:

 >4.F.1.1.5
MDSIRPATFQIPAAVRELGWAALLLFFVLLSVHEWFSPPGWFGLLAILIFATQGALILTR
WPARQNFGWANRTTLLRSILVVSLVAWAPFLPAADSSALWIYGVACLIALILDGVDGKVA
 >2.A.4.2.8
MSPSRTARLYFLLVLDLLFFVLEISIGYAVGSLALVADSFHMLNDVVSLIIALYAIKLAA
SSTPTTRYSYGWHRAEILAALVNGVFLLALCFTITLEALERFFSTPEISNPKLIVLVGSL
 >2.A.4.5.2
IASDIRRILHRHGIHSSTIQPEYHPVRDTILEERSKDVNCLISCPPDSACCEVQACCPSY
AGT

only

 >+second_id+\n+sequence

[5heikki]

1. No greater-than sign means that it's not fasta
2. The format you request looks very random in your example

[WouterDeCoster]

The greater-than sign gets auto-formatted I think, so I guess the post doesn't reflect what OP had in mind.

Answer 1

[Daniel]

To code golf-ify the answer, you could do it in fewer keystrokes with sed:

# 21 Keystrokes (+infile.fa)
sed -i 's/^>.\+ />/g' infile.fa

EDIT: Golfing harder:

# 19 Keystrokes (+infile.fa)
sed -i 's/>.\+ />/' infile.fa

Answer 2

[venu]

If I understand it properly, something like this should work.

cat file.fa | paste - - | awk '{print ">"$2"\n"$3}' > new_file.fa

PS: When I copy paste your sequence, there is a gap in the sequence. If it is a formatting problem, it is ok, if not make sure nothing is going wrong.

After reformatting (by genomax2), first linearize the fasta file and use the above.