In my all R packages, I want to find out which packages are installed by biocLite(), how could I do that ? Thank you.

How about this, using the sessioninfo package:

library(sessioninfo)
library(dplyr)

## list all installed packages
pkgs <- installed.packages()[,'Package']

## read description files & parse using sessioninfo
desc <- lapply(pkgs, utils::packageDescription)
source <- vapply(desc, sessioninfo:::pkg_source, character(1))

## combine and filter for BioC only
bioc_pkgs <- data_frame(pkgs, source) %>%
    filter(source == "Bioconductor")

Output looks like:

> bioc_pkgs
# A tibble: 110 x 2
   pkgs             source      
   <chr>            <chr>       
 1 affy             Bioconductor
 2 affyio           Bioconductor
 3 airway           Bioconductor
 4 annotate         Bioconductor
 5 AnnotationDbi    Bioconductor
 6 AnnotationFilter Bioconductor
 7 AnnotationHub    Bioconductor
 8 Biobase          Bioconductor
 9 BiocGenerics     Bioconductor
10 BiocInstaller    Bioconductor
# ... with 100 more rows

Hi ishenweiyan,

this is part of a script I use to document every now and then the packages I have stored by major R version change

# using package_info for the extra info delivered
### -> you might need to change the lib.loc according to system/version
allpkgs <- (devtools:::package_info(rownames(installed.packages(lib.loc = "/Library/Frameworks/R.framework/Versions/3.5/Resources/library")),include_base = T))
allpkgs[grepl("Bioconductor",allpkgs$source),]

# in a similar fashion, for cran/CRAN packages, and leading edge packages you might have got from Github repos
allpkgs[grepl("cran",allpkgs$source,ignore.case = TRUE),]
allpkgs[grepl("Github",allpkgs$source),]

There might be other ways to do it, but I found this to be relatively robust. If others have alternative solutions, I'd be also happy to check them out ;)

A base R solution. These are the repositories that biocLite uses, filtered to exclude the CRAN repository

repos = BiocInstaller::biocinstallRepos()
repos = repos[startsWith(names(repos), "BioC")]

and the packages installed with biocLite

installed_pkgs = rownames(installed.packages())
available_pkgs = rownames(available.packages(repos = repos))
intersect(available_pkgs, installed_pkgs)