Seeking Assistance for Interpreting a Network Representation on
Entering edit mode
3 months ago
Biomed-jeh • 0

Dear BioStars Community,

I am currently seeking guidance and insights regarding the interpretation of the network representation displayed in the attached image sourced from In this visualization, each data point is presumed to correspond to a gene, ultimately translatable into a protein. However, I am interested in understanding the distances between these data points.

Upon closer examination, I've noticed that there are highlighted points that become apparent when interacting with different enrichment aspects. These highlights exhibit varying patterns, with some being closely clustered, while others exhibit a more dispersed arrangement.

I would greatly appreciate it if someone could kindly assist me in deciphering how to interpret and comprehend the underlying meaning of this network representation. Any guidance or insights would be immensely valuable to my bioinformatics research.

Thank you in advance for your expertise and assistance.

Full proteome network

gene-ontology GO enrichment Proteome stringdb • 454 views
Entering edit mode

What is your network? I.e. where is the input gene list coming from?

What I see is a large network of proteins that are known to interact with each other. I may misremember, but I don't think literal distances between nodes takes on any specific meaning with string-db. Instead, the connection of two nodes means there is evidence of their interaction, string-db lists and allows filtering of what evidence is allowed (e.g. co-mention in literature, co-expression, experimental).

To get anything meaningful from this network analysis, I would think you need more specific questions depending on the context of the gene list.

Regarding the highlighted genes in a pathway being more clustered or dispersed, I think this is more a function for easiest display whether they have direct (node-to-node) evidence with each other and how many other nodes they connect to. So in general, nodes within clusters would be more interactive among each other, but that doesn't mean they are more unrelated to any other connected pathway.

Entering edit mode

Thank you for your response.

I have utilized a gene list comprising gene symbols and their corresponding statistical values, obtained through the DESeq2 analysis. These data points have been input into the STRING-DB function, which can be accessed via the following link: You will find sample examples on the page to facilitate your experimentation.

It is worth noting that I do not assign any significant interpretative value to the resulting figure. My interest primarily lies in understanding the underlying purpose of this visualization. It appears to be an endeavor to depict potential groupings of proteins or elucidate their interrelationships. However, I seek clarity on the precise nature of these groupings and relationships.

Entering edit mode

I think the website does a good job of describing what is shown and giving many options to try and interpret the networks.

However, in general, the only thing that really is informative is if the nodes connect or not. If they connect then that means it detects an association between the proteins or genes. If they don't, then that would be considered unknown interaction. There is no more information presented int he visualization than this. It is not like a dimensional reduction plot or any sort of clustering algorithm.

Really the idea is to observe whether proteins are known to be interconnected or not.


Login before adding your answer.

Traffic: 2069 users visited in the last hour
Help About
Access RSS

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6