I have already been able to do ->

Remove the extra informations i don't want :

cut -f1,4,5,9 -d$'\t'  file.gff | sed 's/ID.*Target=//g' | sed 's/-.*\|+.*//g'  > simplified.gff

Sort the simplified file depending on the Genome ref scaffold and remove headers :

sort -k4 simplified_gff > final.gff

sed 's/#!gff//g' final_gff.gff | sed 's/##gff//g' | sed 's/#!processor NCBI annotwriter//g' >  MyGFF.gff

sed -i '/^$/d'  MyGFF.gff

Then i seperated every informations in separated files :

while read line ; do echo "$line" | cut -f1 ; done < MyGFF.gff >> line1                       #Name of Genome 1 contig

while read line ; do echo "$line" | cut -f2 ; done < MyGFF.gff >> line2                       #Start pos of genome 1 contig

while read line ; do echo "$line" | cut -f3 ; done < MyGFF.gff >> line3                       #End pos of genome 1 contig

while read line ; do echo "$line" | cut -f4 ; done < MyGFF.gff >> ligneINTER

while read line ; do echo "$line" | cut -f1 -d " " ; done < ligneINTER >> line4               #Name of GenomeRef scaffold

while read line ; do echo "$line" | cut -f2 -d " " ; done < ligneINTER >> line5               #Start pos of GenomeRef scaffold

while read line ; do echo "$line" | cut -f3 -d " " ; done < ligneINTER >> line6               #End pos of GenomeRef scaffold

Then for each line, i can make a samtools to catch the sequence of my Genome1 and add it to a file that have the GenomeRef scaffold as name :

####wc -l file.gff = X
####END = X

for i in $(seq 1 $END) ; do samtools faidx Genome1.fasta `sed "${i}q;d" line1`:`sed "${i}q;d" line2`-`sed "${i}q;d" line3` >> `sed "${i}q;d" line4`.fasta ;done

But i really don't see how i can add the N's ....