sleuth_prep fail, read_kallisto , is(path, "character") is not TRUE

[DingJingZhi]

new env
conda install --channel bioconda r-sleuth
Total: 245.0 MB
r-sleuth-0.30.0
python-3.11.5
r-base-3.6.1
...
R
library('sleuth')
samples = c("CK1","CK2","CK3","Y22_3HM1","Y22_3HM2","Y22_3HM3")
s2c <- data.frame(sample = samples, condition = rep(c("control", "case"), each = 3), path = paste("kallisto", samples, sep = "/"))

s2c
sample condition path
1 CK1 control kallisto/CK1
2 CK2 control kallisto/CK2
3 CK3 control kallisto/CK3
4 Y22_3HM1 case kallisto/Y22_3HM1
5 Y22_3HM2 case kallisto/Y22_3HM2
6 Y22_3HM3 case kallisto/Y22_3HM3

so <- sleuth_prep(s2c, extra_bootstrap_summary = TRUE)
reading in kallisto results
dropping unused factor levels
.Error in read_kallisto(path, read_bootstrap = FALSE, max_bootstrap = max_bootstrap) :
is(path, "character") is not TRUE

read_kallisto("./kallisto/Y22_3HM3", read_bootstrap = FALSE, max_bootstrap = 100)
result:
kallisto object
transcripts: 4279
original number of transcripts: 4279
Original or Subset: Subsetbootstraps: 0
This read_kallisto("./kallisto/Y22_3HM3", read_bootstrap = FALSE, max_bootstrap = 100) can run !

why sleuth_prep(s2c, extra_bootstrap_summary = TRUE) fail?
How to do ?

[DingJingZhi]

should add in frame : stringsAsFactors = FALSE
then not factor, but keep char!

s2c <- data.frame(sample = samples,
condition = rep(c("control", "case"), each = 3),
path = paste("kallisto", samples, sep = "/"),
stringsAsFactors = FALSE

             )

so <- sleuth_prep(s2c, extra_bootstrap_summary = TRUE)

normalizing est_counts
4135 targets passed the filter
normalizing tpm
merging in metadata
summarizing bootstraps
......
Error in dyn.load(file, DLLpath = DLLpath, ...)
------another ERROR

[DingJingZhi]

Error in dyn.load(file, DLLpath = DLLpath, ...) :
unable to load shared object '/home/djz/miniconda3/envs/rna/lib/R/library/stringi/libs/stringi.so':
libicui18n.so.58: cannot open shared object file: No such file or directory

file:///home/djz/miniconda3/pkgs/icu-58.2-he6710b0_3/lib here found libicui18n.so.58
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/home/djz/miniconda3/pkgs/icu-58.2-he6710b0_3/lib
see r-lib/xml2#219
Ehm, fixed the issue by adding the anaconda lib path to the system ld path:

locate libicui18n.so.58
#/opt/anaconda2/envs/python3/lib/libicui18n.so.58
#/opt/anaconda2/envs/python3/lib/libicui18n.so.58.2
#/opt/anaconda2/lib/libicui18n.so.58
#/opt/anaconda2/lib/libicui18n.so.58.2
#/opt/anaconda2/pkgs/icu-58.2-h9c2bf20_1/lib/libicui18n.so.58
#/opt/anaconda2/pkgs/icu-58.2-h9c2bf20_1/lib/libicui18n.so.58.2
then,
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/home/djz/miniconda3/pkgs/icu-58.2-he6710b0_3/lib

NICE！ all pass