Skip to content
Snippets Groups Projects
Commit 704bfb3b authored by Zhiyu Zhao's avatar Zhiyu Zhao
Browse files

V1.7.1: first commit

parent 6bceed35
No related merge requests found
This diff is collapsed.
To run this analysis tool, first make a copy of the data template, read the instructions in the template and make necessary changes.
To run the tool on BioHPC @ UTSW: Open a BioHPC web visualization node and run the following from command line:
sh /project/CRI/shared/Metabolomics/run_analysis.sh input_path/input_file_name.xlsm output_path/output_file_name.xlsx optional_BioHPC_queue_name #input_file is the copy of the data template with your data. output_file is the analysis result.
#########################################################################
Change Log:
#v1.7.1, 12/04/2019: Figures were saved as both .ps and .png files. Euler plot was disabled for >6 samples. Samples / features can be clustered or not on the heatmaps.
#v1.7.0, 10/22/2019: Figures were saved as .ps files in a Figure folder. Correlation plot was genrated with the old method for better speed. Euler plot was disabled for >10 samples. Y axis of volcano plot could be raw or adjusted p-values. DESeq was renamed RLE as I would implement the DESeq differential expression analysis in the future. RLE is the same method used by DESeq but I slightly modified it to allow for zeros, missing values and / or values below the value filter. Due to this reason I couldn't use the DESeq normalization directly.
#v1.6.9, 10/4/2019: A bug regarding cpglm was fixed. The bug caused an error in cpglm fitting but the class(fit) was still 'cpglm' and fit$converged was still TRUE. Visualization option was added. Version number checkup was added for the data template. Multiple other minor improvements.
#v1.6.8, 09/19/2019: A bug regarding featureDescription was fixed. The bug was present only when data format was "table with feature description" and samples were excluded which happened to cause some features #v1.6.7, 08/29/2019: Correlation plot was enchanced. Euler diagram was added. Software version number was reported in the result file.
#v1.6.6, 08/28/2019: Data table with feature description column was enabled, minor bugs fixed, and volcano plot colors customized.
#v1.6.5, 08/27/2019: A bug regarding number of features per sample when data imputation = "NA" was fixed.
#v1.6.4, 08/08/2019: A bug regarding groupMeanLowerBound was fixed. Definition and default value of groupCountFilter were modified. Data imputation was implemented. Missing values were allowed.
#v1.6.3, 08/05/2019: a bug regarding nSamples and formula string generation (only when technical replicates are present) was fixed.
#v1.6.2, 07/24/2019: a bug regarding normalization size factors was fixed. The bug caused size factors to be applied to incorrect samples. If you have used versions before this fix, please rerun your analysis using this fixed version.
#v1.6.1, 07/03/2019: a bug regarding numerical batch IDs was fixed.
#v1.6, 06/24/2019: Multiple normalization methods; visualization plots; sample exclusion; feature exclusion; cpglm and glm for statistical tests;
#v1.5, 03/13/2019: If logTransform enabled, data=log2(data+logOffset) before statistical tests; p-values are FDR-adjusted; Bugs fixed; normalization enabled;
#v1.4, 01/18/2019: Added functionality. (1) Averaging technical replicates. (2) Auto-IDs.
#v1.1-1.3: Bugs fixed.
#v1.0: Basic analysis including counting features and t-tests.
#########################################################################
File added
#!/bin/bash
if [[ $# < 2 ]]
then
echo "Usage:"
echo "sh this_file.sh input_path/input_file_name.xlsx output_path/output_file_name.xlsx optional_BioHPC_queue_name"
exit
fi
module load R/3.5.1-gccmkl
DIR=`dirname $0`
INPUT_FILE=$1
OUTPUT_FILE=$2
QUEUE=$3
if [ "$QUEUE" == "" ]
then
QUEUE=256GB
fi
srun -p $QUEUE Rscript $DIR/ODA_1.7.1.R $INPUT_FILE $OUTPUT_FILE
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment