About integrin genes from the output

[Pedramto89]

Hi there

I have the results from the CPDB and in several rows within the genes columns, there are 'integrin'. I'd like to know which integrin gene is that? As you know, there are several integrin genes like ITGA1, ITGA2, etc. This is so important for me to evaluate the enrichment of the genes involved in the analysis between cells.

Best

[cakirb]

Hi @Pedramto89,

Could you show us how your results look like? Then we can better provide a way to find the genes that you're interested in. You can also check CellPhoneDB data and search for integrin as I mentioned similarly in #187.

Best, Batu

[Pedramto89]

This is the result I got from the analysis. In the column D, it shows that the 'integrin' is part of the interaction. But no gene symbol has been put for the interaction.

Furthermore, I tried to visualize the results based on the number of interactions and the interaction strength and you can see the figures. Now, I am wondering, what does it mean biologically and conceptually to have negative scores for the interaction strength plot?

Best, Pedram

[cakirb]

Hi @Pedramto89,

"complex_input" of CellPhoneDB data has the information about complexes including integrins, they can be found by searching from the CellPhoneDB page. I've also extracted integrins from "complex_input" in this file: v5.0.0_complex_input_integrin.xlsx

Here you can find the parts of the complexes, however they are stored as UniPort IDs. You can check UniProt website for which genes each UniProt ID corresponds to. For instance: "integrin_a10b1_complex" has P05556 and O75578.

Hope this helps you!!

Best, Batu

[cakirb]

Regarding to the visualisations, I assume the scores were normalised and that should be why you probably have the negative values. If this is the case, negative scores can be interpreted as less strong interaction compared to general.

Best, Batu

[Pedramto89]

Hi @Pedramto89,

"complex_input" of CellPhoneDB data has the information about complexes including integrins, they can be found by searching from the CellPhoneDB page. I've also extracted integrins from "complex_input" in this file: v5.0.0_complex_input_integrin.xlsx

Here you can find the parts of the complexes, however they are stored as UniPort IDs. You can check UniProt website for which genes each UniProt ID corresponds to. For instance: "integrin_a10b1_complex" has P05556 and O75578.

Hope this helps you!!

Best, Batu

Thank you Batu for your prompt response. In the meantime you responded, I was trying to go over the interactions. What I understood from your response is that I should look up every single complex in the Uniprot e.g., and then see which integrin gene has been involved? Is not a time-consuming procedure? As an example, I searched for an interaction first in the CPDB and then I got the following plot. Now, How can I get the integrin gene part? It just shows the other partner which is here 'COL5A1':

[cakirb]

Hi @Pedramto89,

gene_input.csv from CellPhoneDB data has the information of gene names and their UniProt IDs, so you can concatenate complex_input.csv and gene_input.csv by the UniProt IDs to find the gene names of the proteins in the complexes with integrins as alternative to checking them UniProt website one by one. We will consider to add the gene names in complex_input.csv for future releases.

Best, Batu

[Pedramto89]

Thanks @cakirb for your response. I am still dig into the results. I would like to know which partner is receptor and which partner is ligand. In the results, I can see there are two columns. One is receptor_a and the other is receptor_b. In some cases which sounds weird to me is that for both partners it is FALSE or for both it is TRUE. Would you explain what does it mean?

[cakirb]

Hi @Pedramto89,

It is possible to see the cases where the receptor_a and receptor_b columns are both TRUE or both FALSE. For instance, both receptor_a and receptor_b can be FALSE if they are extracellular matrix proteins, and both can be TRUE when both can act as receptors and trigger responses in the corresponding cells (e.g. HBEGF-ERBB2 -- note HBEGF is transmembrane).

Hope this helps!

Best, Batu