Avoid assigning feature truncation and elongation to non-cDNA regions

[nuno-agostinho]

Fixes Ensembl/ensembl-vep#1480: assigning feature_truncation/elongation to SV deletions/insertions in introns is not ideal given that these consequences will be highlighted as HIGH impact even without any obvious functional consequences.

This bug fix simply checks if the SV is located within the cDNA before assigning the consequence.

Testing

Test VEP with the following variants:

• Breakends (such as bnd1 in the example below) should return feature_truncation/elongation
• Insertions/deletions that affect part of the exon (ins/del_in_exon below) should return feature_truncation/elongation
• Insertions/deletions that affect the intron only (ins/del_in_intron below) should NOT return feature_truncation/elongation

1       7936271 bnd1    N       N[12:58877476[,N[8:56445865[    80      .      SVTYPE=BND
1     202347552 del_in_exon    .       <DEL>                           .       .      END=202392047
1     202391650 del_in_intron   .       <DEL>                           .       .      END=202392047
1     202347552 ins_in_exon    .       <INS>                           .       .      END=202392047
1     202391650 ins_in_intron   .       <INS>                           .       .      END=202392047

The following CNVs should still return feature_truncation/elongation consequences for most cases, except a few cases where the affected regions are only introns (see Ensembl/ensembl-vep#1478 to use these IDs with VEP):

nsv4517655
nsv5594649
nsv4517661
nsv4033399
nsv6310981
esv1812726
esv1815690
esv1850194

[EugeneEA]

Hi, is there any progress on that issue? @nuno-agostinho

[likhitha-surapaneni]

LGTM, thanks @nuno-agostinho