Can I make GFF3 from either GFF or genome fasta file?

Question

[jaqx008]

Hello all, I am trying to create a densitymap for a TE.bed file I have. I am using chicken-repeats.inra.fr densitymap GUI ( I dont know the details of how it works). the required input file format is gff3 and I have a GFF that looks like this

Bf_V2_1     1   797 -   bf_rep_71           Unknown         Bf_V2_1  
    848 936 +   (TA)n               Simple_repeat   Bf_V2_1  
    1236    1369    -   CR1-11_BF           LINE/CR1        Bf_V2_1  
    2151    2171    +   (TA)n               Simple_repeat   Bf_V2_1  
    2351    3238    -   bf_rep_71           Unknown         Bf_V2_1  
    3229    3413    +   DNA-X-4_BF          DNA/Unknown     Bf_V2_1  
    3400    3506    +   Harbinger-N11_BF    DNA/Harbinger

Is there a way to convert this to GFF3? or can I make GFF3 from fasta? And is there another way to create the densitymap to show locations of my transposable elements in the genome? I have read suggestions to similar question and non has been very helpful. Thanks

[Sej Modha]

Is there a way to convert this to GFF3?

Are you asking if you can convert a .bed file to .gff file?

or can I make GFF3 from fasta?

No, it is not possible to convert/generate a GFF3 file from fasta file. GFF file usually stores annotation data whereas fasta file contains sequences.

[jaqx008]

by is there a way to convert this to GFF3 I mean convert my GFF to GFF3.

Answer 1

[Beuss]

I guess the format of example has been broken, so I suppose that your input file is like this:

Bf_V2_1 1   797 -   bf_rep_71   Unknown
Bf_V2_1 848 936 +   (TA)n   Simple_repeat
Bf_V2_1 1236    1369    -   CR1-11_BF   LINE/CR1
Bf_V2_1 2151    2171    +   (TA)n   Simple_repeat
Bf_V2_1 2351    3238    -   bf_rep_71   Unknown
Bf_V2_1 3229    3413    +   DNA-X-4_BF  DNA/Unknown
Bf_V2_1 3400    3506    +   Harbinger-N11_BF    DNA/Harbinger

If the positions on a base 1 like gff, use this perl one liner :

perl -nae 'print "$F[0]\tmySource\t$F[4]\t$F[1]\t$F[2]\.\t$F[3]\tRepeatFamily=$F[5]\n"' TE.bed

Output:

Bf_V2_1 mySource    bf_rep_71   1   797.    -   RepeatFamily=Unknown
Bf_V2_1 mySource    (TA)n   848 936.    +   RepeatFamily=Simple_repeat
Bf_V2_1 mySource    CR1-11_BF   1236    1369.   -   RepeatFamily=LINE/CR1
Bf_V2_1 mySource    (TA)n   2151    2171.   +   RepeatFamily=Simple_repeat
Bf_V2_1 mySource    bf_rep_71   2351    3238.   -   RepeatFamily=Unknown
Bf_V2_1 mySource    DNA-X-4_BF  3229    3413.   +   RepeatFamily=DNA/Unknown
Bf_V2_1 mySource    Harbinger-N11_BF    3400    3506.   +   RepeatFamily=DNA/Harbinger

If the positions on a base 0 like bed, use this perl one liner :

perl -nae 'print "$F[0]\tmySource\t$F[4]\t".($F[1] + 1)."\t$F[2]\.\t$F[3]\tRepeatFamily=$F[5]\n"' TE.bed

Output:

Bf_V2_1 mySource    bf_rep_71   2   797.    -   RepeatFamily=Unknown
Bf_V2_1 mySource    (TA)n   849 936.    +   RepeatFamily=Simple_repeat
Bf_V2_1 mySource    CR1-11_BF   1237    1369.   -   RepeatFamily=LINE/CR1
Bf_V2_1 mySource    (TA)n   2152    2171.   +   RepeatFamily=Simple_repeat
Bf_V2_1 mySource    bf_rep_71   2352    3238.   -   RepeatFamily=Unknown
Bf_V2_1 mySource    DNA-X-4_BF  3230    3413.   +   RepeatFamily=DNA/Unknown
Bf_V2_1 mySource    Harbinger-N11_BF    3401    3506.   +   RepeatFamily=DNA/Harbinger

[Carambakaracho]

while I certainly like the perl one liner solution, there's a few problems with the gff3 like format. Col3 is better represented as something like a CDS, mRNA or similar higher order class and $F[4] better goes to the end and there must not be a period following the coordinate

perl -nae 'print "$F[0]\tmySource\tCDS\t$F[1]\t$F[2]\t$F[3]\tName=$F[4];RepeatFamily=$F[5]\n"' TE.bed

See gff specifications for reference https://github.com/The-Sequence-Ontology/Specifications/blob/master/gff3.md

[jaqx008]

Sorry for the late response. I am trying to this now, will post results soon. Thanks