Self edge formation in visualization method

[horaciobam]

Self edge formation in visualization method

[horaciobam]

In the Retroviridae family this problem is very prevalent

[horaciobam]

Examples:

1. Cluster 2: In HIV-1 genome NC_001802, Gag-Pol (335 to 1637) and Pr55(Gag) (335 to 1838) overlap. Same thing happen on Mouse mammary tumor virus NC_001503, where Pr160 (gene = gag-pol-pro) and Pr77 (gene = gag) overlap.

2. Cluster 1: Snakehead retrovirus NC_001724 is an interesting case where a lot of proteins are hypothetical. For example hypothetical protein from 9345 to 9535 overlaps with hypothetical protein from 9345 to 9357. When exploring the genome, there are 17 hypothetical proteins, several of which share similar ORFS.

3. Cluster 3: In Moloney murine sarcoma virus NC_001502, the pol polyprotein fragment coordinates are: 2485..2928; 2927..3388. We are treating every segment of a protein as an individual unit; therefore this protein seems to be overlapping to itself (this is probably a Ribosomal shift on the reading frame).

4. Cluster 5: In Equine infectious anemia virus NC_001450 there is only one hypothetical protein with coordinates 7295..7951 that apparently has a self edge. This one I don't understand why is being created.

Overall, the self-edge count for overlapping genes in the Retroviridae family is the following:

Cluster: 1, Numer of self edges: 29
Cluster: 2, Numer of self edges: 65
Cluster: 3, Numer of self edges: 1
Cluster: 4, Numer of self edges: 34
Cluster: 5, Numer of self edges: 1
Cluster: 6, Numer of self edges: 16

From exploring the situation, it seems that self edges are formed when two proteins of the same genome are very similar to each other, therefore classified on the same cluster. However, we need to fix the current implementation where two parts of the same protein are being interpreted as individual units.

[horaciobam]

Word Clouds for Retroviridae family

[horaciobam]

Post cluster result

[horaciobam]

Cluster plot:

Genome plot