different result between rseg(0.4.4) and rseg(0.4.8RC)

[Abdulrahman Salhab]

Hi,

I used rseg(0.4.4) to identify the enrichment regions for H3K27me3 and H3K9me3 marks. It worked quit well and the results look very nice; I can see large blocks of the diffusive marks. But, when I used rseg(0.4.8RC) on the same data I got really different results; I got very small regions beside the large blocks (in H3K9me3) and useless region in H3K7me3.  I used the default parameters for each version. 

I wanted to come back again to rseg(0.4.4), but when I tried to install it, it failed because of these errors:

make[1]: Entering directory `rseg-0.4.4/src'

make[2]: Entering directory `/rseg-0.4.4/src/common'

g++ -Wall -fPIC -fmessage-length=50  -O3 -I rseg-0.4.4/src/common -I rseg-0.4.4/src/gsl -I rseg-0.4.4/src/smithlib -c  -o Distro.o Distro.cpp

In file included from Distro.cpp:22:

Distro.hpp:30:25: error: gsl/gsl_rng.h: No such file or directory

Distro.cpp:35:29: error: gsl/gsl_randist.h: No such file or directory

Distro.cpp:37:30: error: gsl/gsl_sf_gamma.h: No such file or directory

Distro.cpp:38:29: error: gsl/gsl_sf_zeta.h: No such file or directory

Distro.cpp:39:26: error: gsl/gsl_diff.h: No such file or directory

Distro.cpp:40:39: error: gsl/gsl_statistics_double.h: No such file or directory

.......

.......

.......

I'm not sure if the all necessary files are included in v0.4.4 on your download page http://smithlabresearch.org/downloads/rseg-0.4.4.tar.gz . I couldn't find "gsl" folder. could you please check this error?

and why do the old version work better than the new one ?  

I attached a snapshot, you can see the difference between the two versions

best,

Abdulrahman  

[Song, Qiang]

Hi Abdulrahman,

Thanks for your interest in RSEG. RSEG v0.4.4 does not includes gsl by
default. You need to install GSL separately
(https://www.gnu.org/software/gsl/).  In version 0.4.8, we include gsl
to make it easier to install.

The result from your analysis looks surprising to me because I
indeed calibrated the performance of these two programs previously.
I just quickly go over the default settings in these two versions, and
find that the new version uses Viterbi decoding while the old version
uses posterior decoding. Can you try running the new version with the
-P option? I am not sure whether this will (or should) make any real
difference.

Also, can you post the running log from both versions? It will be very
useful.

Best,
Song Qiang

--
You received this message because you are subscribed to the Google Groups "RSEG Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rseg-support...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Abdulrahman Salhab]

Hi Song,

thanks for your reply. I will do what you suggested 

best 

Abdulrahman