FusionSeq List of programs

From GersteinInfo

Revision as of 17:35, 19 August 2010 by Asboner (Talk | contribs)
Jump to: navigation, search
FusionSeq main web page
User documentation main


Data formats

FusionSeq use a few data formats to perform its operations.

Mapped Read Format (MRF)

This format is defined in the context of RSEQtools. More details can be found here.

Gene Fusion Report (GFR)

This file format defines the relevant information for each fusion transcript candidate. We designed a standard format for the list of fusion candidates called Gene Fusion Report (GFR) The rationale is that different filters can be applied to exclude “false positives” artificial fusions starting from an initial set. We also provide a parser that interprets this format allowing the user to propagate easily any changes to this format. For a given fusion candidate, involving gene A and gene B, the basic GFR format requires the following fields:

  1. ID: the ID of the fusion candidate: typically it contains the sample name and a unique number separated by an underscore. The number is padded with zeros for consistency;
  2. SPER, DASPER and RESPER: see main text for their definition;
  3. Number of inter-transcript reads, i.e. the number of pairs having the ends mapped to the two genes;
  4. P-value of the insert size distribution analysis for the fusion transcript. Since we do not know the actual composition of the fusion transcript, we computed the p-value for both directions: AB (where gene A is upstream of gene B) and BA (where gene B is upstream of gene A);
  5. Number of intra-transcript reads for gene A and gene B, respectively, i.e the number of pairs where both ends map to the same gene;
  6. The type of the fusion: cis, when both genes are on the same chromosome, or trans, otherwise;
  7. Name(s) of the transcripts: all the UCSC gene IDs of the isoforms of each gene in the annotation separated by the pipe symbol '|';
  8. Chromosome of the genes;
  9. Strand information;
  10. Start and end coordinates of the longest transcript for both genes;
  11. Number of exons in the composite model for both genes;
  12. Coordinates of the exons in the composite model: each exon is separated by the pipe symbol '|' and start and end coordinates are comma-separated;
  13. Exon-pair count: it describes which exons are connected and the number of inter-reads;
  14. Inter-reads: the exon and the coordinates of the reads that join the two genes. Exon number, start and end coordinates are reported as comma-separated, with the pipe symbol '|' separating the different pairs;
  15. Reads of the transcripts: the actual sequence of all the inter-reads.

The GFR format can include additional optional information computed in the subsequent processing. For example, it is possible to add gene symbols and descriptions from the UCSC knownGene annotation set.

Breakpoint data format (BP)

Similarly to GFR, the junction-sequence identifier uses a standard format to capture the results of this analysis. For each tile that has at least 1 read aligned to, it reports, comma-separated:

  1. chromosome, start and end coordinates of the first tile, using UCSC notation: “chr:start-end”, although the intervals are 1-based and closed;
  2. chromosome, start and end coordinates of the second tile
  3. All the sequences of the reads mapped to that tile with the offset information, separated by the pipe symbol.

For example, one line may read as:


where two reads support that specific junction.



geneFusions identifies potential fusion transcript candidates from an alignment file.


geneFusions prefix minNumberOfReads 
  • Inputs: Takes an MRF file from STDIN
  • Outputs: Reports GFR to STDOUT
  • Required arguments
    • prefix - the main ID of each candidate, i.e. prefix_0001, prefix_0002, etc.
    • minNumberOfReads - the minimum number of reads required to include a candiate
  • Optional arguments
    • none


... to be continued

Personal tools