can not run methylCRF

[jx...@uic.edu]

Hi, there,

I downloaded all the sample data and followed the instructions given above . However, when I run the last line to get CpG methylation level, the program didn't run, the err file shows

"

-mdir not exist!
ls: invalid option -- '.'
Try `ls --help' for more information.
cat: invalid option -- 'i'
Try `cat --help' for more information.
can't open H1ES/cpg.bed: No such file or directory at ./methylCRF.pl line 110.
"

What happened? How can I fix it?

Thanks a lot for your help.

regards,
jt

[michael]

Hi jt,

-mdir specifies the directory containing the model files. Did you download these in step 3?

wget http://methylcrf.wustl.edu/h1es_mdl.tgz
tar -zxf h1es_mdl.tgz

If tar is installed and runs, the files should be extracted into the directory used in the  mcrf command. If you put the model files in another directory you'll need to change  -mdir to match. Tell me if that doesn't fix it.

best,
michael

[Garen Tian]

I cannot run methylCRF even I downloaded h1es_mdl. It reported an error that: -mdir not exist. I dont know whats happened? Could you give me some advise?

在 2017年5月2日星期二 UTC+8下午6:03:12，michael写道：

[garen...@gmail.com]

Hi,

I want to calculate absolute DNA methylation with MeDIP-seq. However, I tried use mm9 data to run methylCRF, it is reported an error that:

Sun Aug 20 22:01:11 2017 format -mdir and h1es_mdl

-mdir not exist!
ls: invalid option -- '.'
Try 'ls --help' for more information.

Sun Aug 20 22:01:11 2017 make avg windows for -mdir and h1es_mdl
0
awk: fatal: cannot open file `H1ES/cpg.bed' for reading (No such file or directory)
10
awk: fatal: cannot open file `H1ES/cpg.bed' for reading (No such file or directory)
100
awk: fatal: cannot open file `H1ES/cpg.bed' for reading (No such file or directory)
1000
awk: fatal: cannot open file `H1ES/cpg.bed' for reading (No such file or directory)
10000
awk: fatal: cannot open file `H1ES/cpg.bed' for reading (No such file or directory)
0
awk: fatal: cannot open file `H1ES/cpg.bed' for reading (No such file or directory)
terminate called after throwing an instance of 'std::ios_base::failure'
what(): basic_filebuf::underflow error reading the file
10
awk: fatal: cannot open file `H1ES/cpg.bed' for reading (No such file or directory)
terminate called after throwing an instance of 'std::ios_base::failure'
what(): basic_filebuf::underflow error reading the file
100
awk: fatal: cannot open file `H1ES/cpg.bed' for reading (No such file or directory)
terminate called after throwing an instance of 'std::ios_base::failure'
what(): basic_filebuf::underflow error reading the file
1000
awk: fatal: cannot open file `H1ES/cpg.bed' for reading (No such file or directory)
terminate called after throwing an instance of 'std::ios_base::failure'
what(): basic_filebuf::underflow error reading the file
10000
awk: fatal: cannot open file `H1ES/cpg.bed' for reading (No such file or directory)
terminate called after throwing an instance of 'std::ios_base::failure'
what(): basic_filebuf::underflow error reading the file
14 26 127 FGSC_rep1.bed_DIP_d0.cnt
0 0 0 FGSC_rep1.bed_MRE_d0.cnt
14 26 127 FGSC_rep1.bed_DIP_d10.cnt
0 0 0 FGSC_rep1.bed_MRE_d10.cnt
14 26 127 FGSC_rep1.bed_DIP_d100.cnt
0 0 0 FGSC_rep1.bed_MRE_d100.cnt
14 26 127 FGSC_rep1.bed_DIP_d1000.cnt
0 0 0 FGSC_rep1.bed_MRE_d1000.cnt
14 26 127 FGSC_rep1.bed_DIP_d10000.cnt
0 0 0 FGSC_rep1.bed_MRE_d10000.cnt
FGSC_rep1.bed_DIPMRE.cnt column numbers
5
6
Sun Aug 20 22:01:12 2017 making DIPMRE.cnt for -mdir and h1es_mdl
FGSC_rep1.bed_DIPMRE.cnt column numbers
5
6

cat: invalid option -- 'i'
Try 'cat --help' for more information.
can't open H1ES/cpg.bed: No such file or directory at ./methylCRF.pl line 110.

I dont know what is the problem. Could you give me some advise?

[michael]

Hi Garen,

Sorry for the slow response, I was traveling.  The error message suggests that whatever you put for mdir doesn't exist. Can you check to make sure what you gave matches what is on the filesystem? If that doesn't help, please repost with the commnad line you used and an ls in the directory you are running it from.

Also, you mention MeDIP.seq but not MRE-seq. Although we used a MeDIP-seq model in the paper to show the dependence of the results on each data type, we didn't pursue developing this mode. Unfortunately, since that mode doesn't have any validation or testing, we don't support it.

best,
michael