This an official replacement for https://bitbucket.org/feeldead/gfold/
Because bitbucket no longer supports mercurial repo and we don't know how to replace the original GFOLD mercurial repo to git repo, we move GFOLD to github.
Note that https://bitbucket.org/feeldead/gfold/ may have updated README file If you cannot download from bitbucket, please try the following link: http://compbio.tongji.edu.cn/~fengjx/GFOLD/download.html
VERSIONS: V1.1.4. When '-norm Count' is specified, the code wrongly treats 'Count' as the specified normalization countants. This bug is corrected. Thanks Crystal Chaw for digging out this bug.
V1.1.3. If annotation files contain comments (startiing with #), skip them. Previously the program would crash.
V1.1.2. Fixed a bug that when condition two contains multiple samples, RPKM for condition two in .diff file is wrong.
V1.1.1. A new output file for job 'diff' is added. In this output, normalized read counts are calculated. If a new parameter '-d' for gene description is provided, this file will also associate gene to gene description.
V1.1.0. A new column for the gene name is added to the output if the gene name information is available in the annotation file. Therefore the output format is different from V1.0.9.
V1.0.9. A bug in calculating RPKM for job 'diff' is corrected. This bug is caused by integer division of total read count over 1000000 where float division should be used. It would slightly overestimate the RPKM if the total number of reads is greater than or equal to 1000000. It would crash the program if the total number of reads is below 1000000. Quick examples are added to gfold help message. User specified normalization constants are allowed.
V1.0.8. Correct a bug of failing to load non-standard SAM format file. Specifically, when the read in the file is unmappable and the corresponding line contains less than 11 columns (such as output of bowtie), previous versions of GFOLD would crash. The bug is fixed. More detection of the incorrect input is added. A typo in the documentation is corrected.
V1.0.7. More explanation about input/output is added in the documentation. A tiny change is made in the source code such that GFOLD can safely accept read counts not generated by job 'count'.
V1.0.6. A more comprehensive installation instruction is added. GTF format is set to default. RPKM is added in the 'count' output.
V1.0.5. GTF format for the gene annotation is supported.
V1.0.4. When no replicate is available, calculate GFOLD accurately instead of using MCMC. This version also calculates RPKM of each gene and removes some useless columns in the output of the previous version.
V1.0.2. The first released version, which generates the results of the paper.
INSTALLATION
-
To compile the program you need to install the latest GNU gsl library (gsl-1.15) first. If you cannot install GSL in the standard library, you need to execute the following commands:
export CXXFLAGS="-g -O3 -I/your/installed/path/include -L/your/installed/path/lib" export LD_LIBRARY_PATH="/your/installed/path/lib:"$LD_LIBRARY_PATH
in which /your/installed/path/ is the path where you install GSL. The first one is for compiling and the second one is for running gfold. It is suggested that the second command is added to the .bashrc (for Linux) or .bash_profile (for MacOS) file in the home directory. If you added these two lines, then execute the following command to reload you config file
source ~/.bash_profile # If you are using MacOS source ~/.bashrc # If you are using Linux
To check whether your setting is correct, you can execute the following command in the terminal
echo $CXXFLAGS
If the output is empty then some steps above are not correctly followed. If the output is not empty, then the output should be a directory. Check the existence of this directory and make sure that there is a 'gsl' folder in this directory.
-
Compile the program by running command "make". If GSL libraray is not set correctly, errors as the following would occur.
g++ -O3 -Wall -lgsl -lgslcblas -g main.cc -o gfold In file included from GeneInfo.hpp:29, from main.cc:24: Utility.hpp:69:36: error: gsl/gsl_statistics_int.h: No such file or directory Utility.hpp:70:28: error: gsl/gsl_vector.h: No such file or directory ...
If it happens, follow step 1 again. If error remains, try the following command:
g++ -O3 -Wall -g main.cc -o gfold -lgsl -lgslcblas -I/your/installed/path/include -L/your/installed/path/lib
-
The executable program is "gfold" with documentation in directory "doc"
conda install gfold -c bioconda
conda install -c conda-forge gsl
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/home/hmd/miniconda3/lib
Add the last line in .bash.rc