R language calculates Spearman correlation coefficient in parallel to speed up the construction of co occurrence network

Altogether （co-occurrence network） Network analysis has increasingly become an important part of microbial ecology
Part of the , It has become a hot technology in the current articles . utilize spearman Correlation analysis is to construct co-occurrence
The important method of network , But because of OTU table There are often thousands of lines , use R Self contained corr.test() Letter
Counting is time-consuming , It seriously restricts our analysis speed . Yes spearman The correlation analysis is parallelized
Running can greatly save computing time , For this we wrote spearman Correlation analysis function to achieve parallel
Standardized operation . For the convenience of explanation , This article takes OTU table Take the data , Yes OTU Go in twos spearman phase
Turn off sex branch Analysis , a have to phase Turn off system Count r and display Writing sex p value . I People take since Oneself hand Write Of Letter Count
network_construct() And psych In bag corr.test() Function both the running time and the calculated results
Compared , Our own function network_construct() The calculation time is far less than corr.test() Function Union
The results are the same , Concrete R The code is shown below

library(psych)
system.time(corr.test(otu[,1:500],use=