Skip to content
This repository has been archived by the owner on Jan 24, 2024. It is now read-only.

What do senders, receivers, mediators and influencers mean? #7

Open
newcounts opened this issue Jul 31, 2020 · 19 comments
Open

What do senders, receivers, mediators and influencers mean? #7

newcounts opened this issue Jul 31, 2020 · 19 comments
Labels
documentation Improvements or additions to documentation

Comments

@newcounts
Copy link

Hi,
The output graph of function netVisual_signalingRole has senders, receivers, mediators and influencers. But I don’t understand the exact meaning of mediators and influencers. How those cell acts as a(n) mediator and/or influencer.
image

thanks.
best wish

@sqjin
Copy link
Owner

sqjin commented Jul 31, 2020

@newcounts Hello, the mediator score is quantified by flow betweenness metric from graph theory. Intuitively, it measures a group of cells’ capability as gatekeeper to control communication flow between any two cell groups. The influencer is quantified by a hybrid measure called information centrality. Basically, for information flow within a signaling network, a higher value indicates greater influence on the information flow. The diagram in Figure 1e is helpful for understanding such measures. We also described these measures in the Method section.

@MMriran
Copy link

MMriran commented Aug 1, 2020

Hi,I had noticed that in your paper you methioned that “the majority of TGFβ interactions among wound cells are paracrine, with only one fibroblast and one myeloid population demonstrating significant autocrine signaling”,how can I get such information from the heatmap (fig2b in your paper)?

@sqjin
Copy link
Owner

sqjin commented Aug 1, 2020

Hi @MMriran, in fig2b, slightly higher scores of both Sender and Receiver for FIB-D suggest that FIB-D is the autocrine signaling. The same case is for MYL-A population.

@gynecoloji
Copy link

@sqjin
Hi Dr. sqjin,

Thanks for your clarification for differences between mediators and influencers. But I still do not understand the real difference between them even if you tell me how you calculate it. Because I am not a mathematician and do not fully understand the metrics used in your calculation. I am sorry about it.
So Could you please clarify them with more obvious words?

Sincerely,

Ji

@sqjin
Copy link
Owner

sqjin commented Sep 11, 2020

Hi @gynecoloji , these measures are indeed similar, and both of them can represent the importance of cell groups in controlling the communication flow. But I think they emphasis on different aspects. In particularly, mediators is quantified by 'betweeness', which has been widely used to indicate the importance of nodes in the network. You might can simply google it and learn more about it (https://en.wikipedia.org/wiki/Betweenness_centrality). From our Figure 1e, you might can find some slightly differences between mediators and influencers. You can also learn about from the references in our paper.

@sqjin sqjin added the documentation Improvements or additions to documentation label Oct 7, 2020
@jv-20
Copy link

jv-20 commented Oct 23, 2020

Hi @sqjin, a quick question in relation to accessing results mediator & influencer scores.

For example, when I do cellchat.obj@netP$centr$ I can access the centrality measures such as outdegree, indegree, hub, authority, eigen, page_rank ,betweeness, flow_between and info.

I am trying to extract the scores for mediators and influencers. In the paper you mentioned flow betweeness is used for identifying mediators. May I ask what is the other score called "betweeness" used for?

Is the value for influencers stored in the cellchat results? In the paper you mentioned "Information centrality score provides a hybrid measure, for example by combining closeness and eigenvector". I see object stored named eigen but not closeness. Could you please advice how to access the value for closeness?

Thank you, apologies if I am asking the obvious. Would appreciate your thoughts.

@sqjin
Copy link
Owner

sqjin commented Oct 26, 2020

Hi @jv-20 , the other score "betweenness" is a classic concept/measure in the graph theory. You can simply search it online.

The influencers can be accessed via centr$info. You can check the documentation of netVisual_signalingRole to see the measures and the corresponding new names. CellChat does not compute the closeness metric.

@jv-20
Copy link

jv-20 commented Oct 26, 2020

Many thanks @sqjin. The documentation is helpful, thank you

@sunny-yangyang
Copy link

Thanks for creating this excellent tool and giving us such detailed explanation. But I'm still confused about the biological roles of Mediator and Influencer. If Sender can be seen as the Ligand, Receiver can be seen as the Receptor, How about the role and function of Mediator and Influencer in cell-cell communication? Does Mediator can be seen as the Cofactor or anything else? Could you please provide an example of the Mediator and Influencer in a specific pathway? Thank you soooo much!

@sunny-yangyang
Copy link

Thanks for creating this excellent tool and giving us such detailed explanation. But I'm still confused about the biological roles of Mediator and Influencer. If Sender can be seen as the Ligand, Receiver can be seen as the Receptor, How about the role and function of Mediator and Influencer in cell-cell communication? Does Mediator can be seen as the Cofactor or anything else? Could you please provide an example of the Mediator and Influencer in a specific pathway? Thank you soooo much!

Sorry to correct my question, If Sender could express Ligand, Receiver could express Receptor, Does Mediator or Influencer express different moleculars such as Agonist/Antagonist OR Agonist/Antagonist Receptor? I just wonder the difference roles between Mediator and Influencer in cell-cell communication pathway. Sorry to interupt you if my questions are silly and still looking forward your reply, Thank you so much~

@sqjin
Copy link
Owner

sqjin commented Dec 12, 2020

Thanks for creating this excellent tool and giving us such detailed explanation. But I'm still confused about the biological roles of Mediator and Influencer. If Sender can be seen as the Ligand, Receiver can be seen as the Receptor, How about the role and function of Mediator and Influencer in cell-cell communication? Does Mediator can be seen as the Cofactor or anything else? Could you please provide an example of the Mediator and Influencer in a specific pathway? Thank you soooo much!

Sorry to correct my question, If Sender could express Ligand, Receiver could express Receptor, Does Mediator or Influencer express different moleculars such as Agonist/Antagonist OR Agonist/Antagonist Receptor? I just wonder the difference roles between Mediator and Influencer in cell-cell communication pathway. Sorry to interupt you if my questions are silly and still looking forward your reply, Thank you so much~

These measures are used to quantify the importance of cell types in certain effect and they do not represent any molecules. Cell types with high sender scores suggests they are the major signaling sources.

@sunny-yangyang
Copy link

sunny-yangyang commented Dec 12, 2020 via email

@sunny-yangyang
Copy link

sunny-yangyang commented Dec 12, 2020 via email

@sqjin
Copy link
Owner

sqjin commented Dec 14, 2020

In addtion, I just wondering if the "contribution of each L-R ligand "(Fig 2c,2e in your bioRxiv preprint 10.1101/2020.07.21.214386) includs all the factors such as Ligand/Receptor and co-factors like Agonist/Antagonist, co-stimulatory/co-inhibitory Receptor?

------------------ 原始邮件 ------------------ 发件人: "sqjin/CellChat" <[email protected]>; 发送时间: 2020年12月12日(星期六) 下午3:57 收件人: "sqjin/CellChat"<[email protected]>; 抄送: "846720602"<[email protected]>;"Comment"<[email protected]>; 主题: Re: [sqjin/CellChat] What do senders, receivers, mediators and influencers mean? (#7) Thanks for creating this excellent tool and giving us such detailed explanation. But I'm still confused about the biological roles of Mediator and Influencer. If Sender can be seen as the Ligand, Receiver can be seen as the Receptor, How about the role and function of Mediator and Influencer in cell-cell communication? Does Mediator can be seen as the Cofactor or anything else? Could you please provide an example of the Mediator and Influencer in a specific pathway? Thank you soooo much! Sorry to correct my question, If Sender could express Ligand, Receiver could express Receptor, Does Mediator or Influencer express different moleculars such as Agonist/Antagonist OR Agonist/Antagonist Receptor? I just wonder the difference roles between Mediator and Influencer in cell-cell communication pathway. Sorry to interupt you if my questions are silly and still looking forward your reply, Thank you so much~ These measures are used to quantify the importance of cell types in certain effect and they do not represent any molecules. Cell types with high sender scores suggests they are the major signaling sources. — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

The contribution of each L-R ligand is calculated based on the overall communication probability (i.e., sum of all the weights) of the inferred network. It is NOT related with any signaling molecule!

@gAleryani
Copy link

gAleryani commented Jan 26, 2022

Hi @sqjin ,
Great tool! And the manual / vignette is very comprehensive, great to look at for the analysis of all kinds of research scenarios. :)

I do think though that some more explanation for how to interpret the data is required, one that is within the context of cellular biology, after all, the tool is called "CellChat". Even after reading this board I still do not fully comprehend what the difference between "mediator" and "influencer" is, this within the practical context of cell-cell interaction of any system like let's say the tumour microenvironment or organ development.

I would really appreciate an explanation! Thanks in advance for your time :)

@sunny-yangyang
Copy link

sunny-yangyang commented Jan 26, 2022 via email

@Lucky-Yangyang
Copy link

Hi@MMriran,in fig2b, slightly higher scores of both Sender and Receiver for FIB-D suggest that FIB-D is the autocrine signaling. The same case is for MYL-A population.
Hi @sqjin ,
In fig2b, the scores of FIB-D in Mediator and Influencer are significantly higher than those in Sender and Receiver, why suggest that FIB-D is the autocrine signaling? Could I comprehend if a cell type have both Sender and Receiver score, even though that is not so high, which is a autocrine signaling? And if a cell type have Sender or Receiver or none of them , even though the score of Sender or Receiver is so high, which still mean that is a paracrine signaling?

@sunny-yangyang
Copy link

sunny-yangyang commented Apr 19, 2023 via email

@sqjin
Copy link
Owner

sqjin commented Apr 24, 2023

Hi@MMriran,in fig2b, slightly higher scores of both Sender and Receiver for FIB-D suggest that FIB-D is the autocrine signaling. The same case is for MYL-A population.
Hi @sqjin ,
In fig2b, the scores of FIB-D in Mediator and Influencer are significantly higher than those in Sender and Receiver, why suggest that FIB-D is the autocrine signaling? Could I comprehend if a cell type have both Sender and Receiver score, even though that is not so high, which is a autocrine signaling? And if a cell type have Sender or Receiver or none of them , even though the score of Sender or Receiver is so high, which still mean that is a paracrine signaling?

autocrine signaling means the signaling among Fibroblasts and this is not related with Mediator/Influencer.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
documentation Improvements or additions to documentation
Projects
None yet
Development

No branches or pull requests

8 participants