Useful tricks and tips for Galaxy users

"You must feel the Force around you." Yoda

• Text processing
• HTS
• Workflows
• More resources
• Disclaimer

Text processing

• Convert comma separated files into tab-separated files
  Convert delimiters to TAB
• FASTA files with unique sequences
  FASTA-to-Tabular → Unique occurrences of each record (advanced parameters) → Tabular-to-FASTA
• Remove sequences with N or any other character
  FASTA-to-Tabular → Filter data on any column using simple expressions with
  (condition: c2.find('N') != -1) → Tabular-to-FASTA
• Extracting the 3rd column from a 5 column file
  Cut columns from a table with c3
• Reorder columns or column swap
  Cut columns from a table with c3,c2,c1
• Count how often one entry appears in column 1
  Datamash with Group by fields: 1 and Operation to perform: count
• Remove all lines that contain a character (comma in this case)
  Text transformation with sed with SED Programm: /,/d
• Group all rows where column 1, 4 and 5 are identical
  Datamash with Group by fields: 1,4,5
• Column-to-rows and rows-to-columns (transpose matrix)
  Transpose rows/columns
• Make your files smaller, e.g. for testing; subsampling of files
  Select random lines from a file
• Make your sequence files smaller, e.g. for testing; subsampling sequences
  Sub-sample sequences files
• Merge two files together according to one column in every file
  Join two files
• Add unique column
  Add column to an existing dataset with iterate: Yes
• Get rid of all rows where column 2 has values greater than 0
  Filter data on any column using simple expressions with c2<=0
• Get all rows where column 4 starts with hsa
  Filter data on any column using simple expressions with c4.startswith('hsa')
• Get rid of all rows where the sum of column 2 and 3 is greater than 10
  Filter data on any column using simple expressions with c2+c3<=10
• Get rid of all rows where the length of my text in column 2 is greater than 10
  Filter data on any column using simple expressions with len(c2)<=10
• Create new rows for every comma separated value in column 3; Unfolding
  Unfold columns from a table with Column 3 and Comma
• Split the first four characters of a line into it's own column
  Replace Text in entire line with Find Pattern: ^(.{4}) and Replace Pattern: &\t
• Add the basepairs "TA" to the end of each sequences
  FASTA to Tabular → Add column with TA → Merge Columns → Cut columns → Tabular to FASTA
• Add a quotation mark to every row
  Compute an expression on every row with chr(34) (34 is the ASCII code for ")
• Count all columns with numbers that do not contain 0. Usefull if you want to calculate the mean but want to exclude all columns that are 0.
  Compute an expression on every row with bool(c1) + bool(c1) + bool(c3) ...
• Calculate log2 (not log10) from a column (e.g. c1) adding a new column Compute an expression on every row with log(c1,2)

HTS

• Map RNA-seq data
  HISAT or TopHat
• Map DNA-seq data
  Bowtie or BWA
• Map methylC-seq data
  Bismark
• Downsample BAM/SAM files
   BAM/SAM Mapping Stats will give you the number of reads/read pair in your BAM file in case you don't know it already. Then you just divide the number of reads you want to downscale to with the number of reads you have and use this fraction as the probability in Picard – Downsample SAM/BAM.
• Get all genes that are covert by reads
  htseq-count with a gene annotation GTF file on your BAM file → Filter data on any column using simple expressions with c2>0
• Extract sequences from intercal files, like gff, bed, gtf. Returning FASTA file →
  Extract Genomic DNA using coordinates from assembled/unassembled genomes

Workflows

• Find two genes located nearby
  Description :: Tool Shed

More Resources

• Galaxy 101 - a must read for all HTS Padawan: https://galaxyproject.org/learn/galaxy-ngs101/
• A lot of videos to learn using Galaxy while eating popcorn: https://vimeo.com/galaxyproject

Disclaimer

All tools mentioned here are available from the Galaxy Tool Shed. Kindly ask your Galaxy Administrator to get access to them.