Two evidences for the same molecular interaction are not aggregated

General
1

In the BMP7 profile page you can see, two different interaction evidences with TRIM27. Why are those separate? Shouldn’t they be aggregated and scored together? If not, why? And if I wish to aggregate them, what is the best way?

image
image1720×512 47.6 KB

image
image1720×561 52.5 KB

2

Hi @Aner_Ottolenghi ,

I agree this is not very clear.

I looked at the interaction dataset to explore how the raw data looked like and the problem is that the interactions are reported against different isoforms of BMP7. One ot the reviewed uniprot accession P18075 and another one to the unreviewed A8K571. This is not reflected in the web UI and we might need to specify not only the isoforms for B, but also the isoform for A.

+--------------+---------------+---------------+------------------+---------------+---------------+------------------+--------------------+--------------------+-----+-------+
|sourceDatabase|        targetA|           intA|intABiologicalRole|        targetB|           intB|intBBiologicalRole|            speciesA|            speciesB|count|scoring|
+--------------+---------------+---------------+------------------+---------------+---------------+------------------+--------------------+--------------------+-----+-------+
|        string|ENSG00000101144|ENSP00000379204|  unspecified role|ENSG00000204713|ENSP00000366404|  unspecified role|{human, Homo sapi...|{human, Homo sapi...|    1|  0.183|
|        string|ENSG00000204713|ENSP00000366404|  unspecified role|ENSG00000101144|ENSP00000379204|  unspecified role|{human, Homo sapi...|{human, Homo sapi...|    1|  0.183|
|        intact|ENSG00000101144|         P18075|  unspecified role|ENSG00000204713|         P14373|  unspecified role|{human, Homo sapi...|{human, Homo sapi...|    4|   0.67|
|        intact|ENSG00000101144|         A8K571|  unspecified role|ENSG00000204713|         P14373|  unspecified role|{human, Homo sapi...|{human, Homo sapi...|    3|   0.56|
|        intact|ENSG00000204713|         P14373|  unspecified role|ENSG00000101144|         A8K571|  unspecified role|{human, Homo sapi...|{human, Homo sapi...|    3|   0.56|
|        intact|ENSG00000204713|         P14373|  unspecified role|ENSG00000101144|         P18075|  unspecified role|{human, Homo sapi...|{human, Homo sapi...|    4|   0.67|
+--------------+---------------+---------------+------------------+---------------+---------------+------------------+--------------------+--------------------+-----+-------+

We will try to address this issue in the next ticket Add isoforms for molecular interactor A to minimise confusion on duplicated rows · Issue #3905 · opentargets/issues · GitHub