博文
A dialogue about single heavy hammer vs. multi-hammer approach
|
Several days ago, I post a paper about Chinese traditional medicine, in which a problem of hammer vs. multi-hammer approach was raised.
I discussed with my colleague in bio-network research. It turns out that it convinced my guess: multi-hammer approach could be more efficient than single heavy hammer approach.
https://blog.sciencenet.cn/blog-3468-27851.html
上一篇：科学与审美
下一篇：Apoptosis, autophagy 和德国的系统生物学会议
I discussed with my colleague in bio-network research. It turns out that it convinced my guess: multi-hammer approach could be more efficient than single heavy hammer approach.
Problem:
For a stable network with robustness, which way has more effects on it: change one route heavily, or change several place slightly?
Formulated question:
Supposing there is a food web, if I want to decrease an animal (for example mouse's) food quantity by 50%, which is more efficient to reach this goal: change one chain in food web by 90%, or change several route by 30%?
Answer:
Food webs and other evolved biological networks almost always tolerate a 90% change of a RANDOMLY picked link, however the same networks disintegrate if a link with high betweenness centrality is INSIGHTFULLY SELECTED. So if you compare to a strong but RANDOM link change then I could imagine that SEVERAL small changes have a higher chance to hit a link of high betweenness centrality for which even a small change has more impact on the whole network than the single, random, large change. For ODE dynamics on networks there is a powerful concept of Metabolic Control Analysis (MCA) to access the impact of link changes.
https://blog.sciencenet.cn/blog-3468-27851.html
上一篇：科学与审美
下一篇：Apoptosis, autophagy 和德国的系统生物学会议
