CLIPSeqTools

Suite to analyse CLIP-Seq datasets

View project onGitHub

Available Toolboxes

A) clipseqtools-preprocess

The CLIPSeqTools toolboxes require that the alignment data are stored in a database (e.g. SQLite, MySQL, etc.). The advantages of following this approach come from the fact that databases support efficient range queries and at the same time allow for straightforward annotation of the reads. The importance of the former is evident - almost all analyses require running some sort of range query on the datasets. The latter, is especially important for CLIP-Seq analysis. The reason is that in certain cases an analysis might need to be limited on a particular subset of the data probably corresponding to a specific annotation (e.g. genic areas) or exclude an annotation (e.g. reads overlapping with repeat elements) or maybe a combination of both cases. Databases allow to do such selections particularly easy with usually no or small additional computational load.

clipseqtools-preprocess is a module that is used to prepare the files (database) that clipseqtools and clipseqtools-compare require. It will get from a FastQ file with CLIP-Seq data to an SQLite database with most critical default annotations included. In more detail, it will process the FastQ file, align the reads on a reference genome, annotate the alignments with genic, repeat masker and evolutionary conservation information and finally prepare an SQLite database compatible with clipseqtools and clipseqtools-compare.

Supported commands

Each command of clipseqtools-preprocess is designed to perform a well defined task. To invoke a command use:

clipseqtools-compare <command>

clipseqtools-preprocess supports the following commands which can run independently or as a predefined pipeline.

  1. all - Will run all of the commands as a pipeline. This is probably the most common option to use unless you need very fine-grained control on what is happening.

  2. cut_adaptor - Remove the adaptor sequence from the 3'end of reads using the cutadapt program.

  3. collapse_fastq - Remove fastq reads that have exactly the same sequence. Can be used to remove possible PCR artifacts.

  4. trim_fastq - Trim a specific number of nucleotides from the start and/or the end of the sequence. Can be used for the removal of randomized barcodes. In conjuction with 'collapse_fastq' they can be used to collapse barcoded reads and then trim the barcode part.

  5. star_alignment - Align the reads on a reference genome using the STAR program.

  6. cleanup_alignment - Sort STAR alignments and keep only a single record for multimappers.

  7. sam_to_sqlite - Load the SAM file with the alignments in an SQLite database.

  8. annotate_with_deletions - Annotate alignments with deletions.

  9. annotate_with_file - Annotate alignments contained within regions from a BED/SAM file.

  10. annotate_with_genic_elements - Annotate alignments with genic information (transcripts, exons, 3'UTRs, etc).

  11. annotate_with_conservation - Annotate alignments with phastCons conservation scores.

Details for database

clipseqtools-preprocess will create a database compatible with clipseqtools. The fields of the database table are extracted directly from the SAM file with the alignments. Not all information are extracted from the SAM file. Only the information that is required by clipseqtools.

Specifically the fields that are included in the database are:

  • id - An autoincrement id.

  • strand - Can be +1 for "+" strand and -1 for "-" strand depending on where the read alinged to.

  • rname - Name of the reference sequence on which the read aligned to. Usually a chromosome name.

  • start - Position on reference sequence where the alignment starts, 0-based inclusive.

  • stop - Position on reference sequence where the alignment stops, 0-based inclusive.

  • copy_number - Number of reads with the same sequence that align on this position.

  • sequence - Sequence of the read.

  • cigar - CIGAR alignment string - see SAM file format documentation for details.

  • mdz - MD:Z alignment tag - see the SAM file format documentation for details.

  • number_of_mappings - Number of alternative places in which the read aligns to.

  • query_length - Length of the read.

  • alignment_length - Length of the alignment - can be different from read length due to insertions or deletions.

Extra columns will be added to the database, if and when the annotation commands run. These annotation columns are:

  • transcript - Defined if the read is contained in a transcript and not defined otherwise.

  • coding_transcript - Defined if the read is contained in a coding transcript and not defined otherwise.

  • noncoding_transcript - Defined if the read is contained in a non coding transcript and not defined otherwise.

  • exon - Defined if the read is contained in an exon and not defined otherwise.

  • utr5 - Defined if the read is contained in a 5'UTR and not defined otherwise.

  • cds - Defined if the read is contained in a coding sequence and not defined otherwise.

  • utr3 - Defined if the read is contained in a 3'UTR and not defined otherwise.

  • rmsk - Defined if the read is contained in a repeat element (Repeat Masker) and not defined otherwise.

  • deletion - Defined if the read alignment has a deletion and not defined otherwise.

  • conservation - Conservation score for the read. The score is calculated as the average phastCons score of all the nucleotides of the read. To minimize storage needs, the phastCons conservation score is multiplied by 1000 to convert it from floating point number to integer.

B) clipseqtools

clipseqtools is the main toolbox of the CLIPSeqTools suite. It runs analyses on a single dataset. It offers a wide selection of tools that cover many aspects of a CLIP-Seq analysis pipeline.

Supported commands

Each command of clipseqtools is designed to perform a well defined task. To invoke a command use:

clipseqtools <command>

clipseqtools supports the following commands which can run independently or as a predefined pipeline.

  1. all Will run all of the commands as a pipeline. This is probably the most common option to use unless you need very fine-grained control on what is happening.

  2. reads_long_gaps_size_distribution Measure the size distribution of long alignment gaps (eg. alignment on exon-exon junctions) produced by a gap aware aligner.

  3. size_distribution Measure the size distribution for reads.

  4. cluster_size_and_score_distribution Assemble reads in clusters and measure their size and number of contained reads distribution.

  5. count_reads_on_genic_elements Count reads on transcripts, genes, exons and introns.

  6. distribution_on_genic_elements Measure how reads are distributed along the length of 5'UTR, CDS and 3'UTR.

  7. distribution_on_introns_exons Measure how reads are distributed along the length of exons and introns.

  8. genome_coverage Measure percent of genome covered by reads.

  9. genomic_distribution Count reads on genes, repeats, exons , introns, 5'UTRs, ...

  10. nmer_enrichment_over_shuffled Measure the enrichment of Nmers within the reads over shuffled reads.

  11. nucleotide_composition Measure the nucleotide composition along reads.

  12. conservation_distribution Measure the number of reads at each conservation level.

Running analysis on subsets of data

Since clipseqtools relies on database tables, the filtering and run of an analysis on subsets of data is particularly straightforward. The only thing a user has to do is give the filtering criteria when executing each of the commands. The syntax for the filtering criteria is easy and intuitive and probably best explained with an example.

Example:

To run an analysis only on reads that are highly conserved, have a deletion and are not repeats, the following flags should be added when running a command:

--filter conservation=">500" --filter deletion="def" --filter rmsk="undef"

The supported operators for creating a filter are: >, >=, <, <=, =, !=, def, undef.

C) clipseqtools-compare

clipseqtools-compare is a toolbox that can be used to compare CLIP-Seq datasets with each other.

Supported commands

clipseqtools-compare supports the following commands which can run independently.

  1. all - Will run all of the commands as a pipeline. This is probably the most common option to use unless you need very fine-grained control on what is happening.

  2. libraries_overlap_stats - Count the reads of library A that overlap those of reference library B.

  3. libraries_relative_read_density - Measure the density of reads of library A around the reads of a reference library B.

  4. compare_counts - Do Upper Quartile normalization upon specified columns (containing counts) of tables.

    Note: This command when called by the all command will compare the counts for genes, transcripts, exons and introns.

  5. join_tables - Combines multiple counts tables into one large table that can be exported.

D) clipseqtools-plot

clipseqtools-plot is an application that can be used to create figures from the output of clipseqtools and clipseqtools-compare.

NOTE: Usually you don't need to run any of these commands because they are automatically called from the corresponding clipseqtools or clipseqtools-compare commands when the --plot flag is given.

Supported commands

clipseqtools-plot supports the following commands which run independently.

  1. cluster_size_and_score_distribution - Create plots for script cluster_size_and_score_distribution.

  2. distribution_on_genic_elements - Create plots for script distribution_on_genic_elements.

  3. distribution_on_introns_exons - Create plots for script distribution_on_introns_exons.

  4. genomic_distribution - Create plots for script genomic_distribution.

  5. libraries_relative_read_density - Create plots for script libraries_relative_read_density.

  6. nucleotide_composition - Create plots for script nucleotide_composition.

  7. reads_long_gaps_size_distribution - Create plots for script reads_long_gaps_size_distribution.

  8. size_distribution - Create plots for script size_distribution.

Need More Details?

The most up-to-date information regarding any of the toolboxes and the commands can be found in the application itself. To find more information for a particular toolbox, just run the toolbox invoking the help command. Alternativelly if you want to see the full manual you can invoke the man command.

Example:

clipseqtools help
clipseqtools man

If you want to get information for a particular command then execute the command with the --help flag. Alternativelly if you want to see the full manual you can invoke the man command followed by the command name.

Example:

clipseqtools genome_coverage --helpcommands
clipseqtools man genome_coverage