stevekm

#!/bin/bash
# I needed a unique random name for files generated by scripts
# so they don't get overwritten on subsequent compiles
# use mktemp to generate a randomly named tmp directory, save the name to a variable
BASE=$(mktemp)
echo $BASE # the output looks like this: /tmp/tmp.XMbgXwwviZ

# we only want the base random string, not the file path
# use these two steps to strip out the file path
BASE2=${BASE///tmp}

stevekm

#!/bin/bash

$PATTERN="MA0007.2_AR" # the gene we are interested in
cd ~/OutputDirectory/ # where your data files are stored

if [[ -n $(grep -Fl $PATTERN *.txt) ]] ; then
# grep; -F fixed pattern, -l ouput file name of matches
# [[ -n ]]; test (see man test), -n nonzero string length
# grep searches files for our pattern in all .txt files,
# and outputs filenames if found

stevekm

# the directories we are using
DIRS<-c("/home/user/DataOutput/",
        "/home/user/DataOutput/A/",
        "/home/user/DataOutput/B/")

# iterate through the vector of directories
for(i in 1:length(DIRS)){
  if(dir.exists(DIRS[i])){
  # delete the directory
    unlink(DIRS[i],recursive = T,force = T)

stevekm

#!bin/bash
for i in {1..5}
do
( echo $'\n Starting primary UP motif analysis in parrallel' $i >> $WorkDir/HOMER_TIMER.txt
  date >> $WorkDir/HOMER_TIMER.txt
sleep $[ ( $RANDOM % 30 )  + 10 ]s
findMotifsGenome.pl $UpARPeaks hg19 $MotifDir/Up/$i -size 200 -p 30
echo $'\n Finished primary UP motif analysis ' $i >> $WorkDir/HOMER_TIMER.txt
date >> $WorkDir/HOMER_TIMER.txt ) &
done

stevekm

#!/bin/bash
find -type f -size +4000M | while read FILE; do 7z a /media/user/StorageDevice/"${FILE}.7z" -mx0 -v4000m "$FILE"; done

stevekm

#!/bin/bash
rsync -avzheR --progress --max-size=500K -e "ssh -p 22" user@server.com:/home/user/ /home/user/ServerBackups/
# "ssh -p 22" = connect via ssh, port 22
# -a, --archive               archive mode; same as -rlptgoD (no -H)
# -v, --verbose               increase verbosity
# -z, --compress              compress file data during the transfer
# -h, --human-readable        output numbers in a human-readable format
# -e, --rsh=COMMAND           specify the remote shell to use
# -R, --relative              use relative path names
# rsync /source/ /destination/

stevekm

#!/bin/bash
# save directory names and their associated ID's into an index log

# set the proper permissions for all files created by this script
umask 007 

# make sure that the log file exists
LOG=/filepath/PARENT_DIR/logs/index_log.tsv
touch $LOG

stevekm

#!/bin/bash
gzip -cvr /path/to/input/dir/ > /path/to/output.gz
# -c std out
# -r recursive through dir
# -v verbose, of course
# pipe to file you want output in 

stevekm

#!/bin/bash
# set the proper permissions for all files created by this script
umask 007 
# make sure that the log file exists
LOG=/path/to/log.tsv
touch $LOG

# find the list of directory names to enter into the log, they all start with a 1
FILES=$(find /path/to/dirs/1* -maxdepth 0 -type d)
for i in $FILES; do

stevekm

#!/bin/bash

INPUT_DIR=/path/to/dir/
OUTPUT_FILES=/path/to/output.tar.gz
OUTPUT_DIR=/path/to/

# create a tar.gz from a directory
tar -cvzf $OUTPUT_FILES $INPUT_DIR

# generate a md5sum for the tar.gz