Hi,

I do not know how you built your STRING analysis but as a suggestion, maybe you can identify which genes seems to be altered inside the enriched pathways that you want to dig into. Then built the STRING analysis using your KO gene and all those genes specifically altered linked to that pathway to see if there is any protein protein interaction. Furthermore, you could also check in REACTOME as there you have more detailed info about the interaction types (activation, inactivation, complexes...) If you are comfortable using Cytoscape a fast exploratory way of looking into this will be running inside Cytoscape clueGo to identify the genes altered inside the pathways of interest. Then REACTOME FI with the list of genes then STRING with the same list of genes. Please note that REACTOME FI have also annotations for proteins that an be activating or repressing the expression fo genes inside of the possible interactions you will get and that, if I understood you correctly is one of the theories you have about your KO gene. Best,

Mar

Thank you Mar Muñiz Moreno for your helpful suggestion.

I built my String Network using all the DE proteins (excluding KO gene).

There is no PPI between KO-gene and genes in overexpressed pathway. Later, I have included all the DE genes and KO-gene and built the network with StringDB (in cytoscape), there is a weak linkage with 3 nodes apart. All intervening nodes are DE proteins.

With reference to Reactome FI, when using all the DEs, my OE genes of interest are in a minor cluster. I will rebuilt with only genes of interest.

I am also surveying the literature and will see how it go. Thank you for your suggestions again.

Best,

Ayaz

Hi Ayaz,

Yes rebuilding the networks using only the DEGs you are interested in pursuing plus your KO gene is a good start. In fact to built a fully connected network STRING/Reactome will need to add some genes in between your genes of interest, instead of disregarding those ones you could try to see the expression values of those genes to see if they were false negative possibly in your differential expression analysis and to see if biologically you find a possible mechanism of dysregulation due to your KO gene linked to the altered functionality of those genes and the connected ones.

Also with REACTOME as is more curated thus the number of interactions you will find is lower if you decrease to much the number of queried seeds you may end with nothing useful. In that case you may try to built first a functionality network, meaning your DEGs of interest plus the first level interactors of those ones, and then query all of them in REACTOME, and again re annotate those first level interactors with the expression levels to see if they could be slightly altered and following the same regulatory sense. Good luck. Kind regards, Mar