lrgasp-submissions

Transcript expression matrix format

The format required for providing quantification results by the user should be a tab-separated matrix in which long read-defined transcripts will be the rows and biological replicates will be columns.

Expression matrix files should be gzip-compressed and named expression.tsv.gz.

Key features

• Header: The first field must be ID, and the rest of the column names should be the replicate labels.
• ID column: Transcripts should have the same IDs as ones provided in GTF, and must be a conforming feature identifier.
• Replicate columns: Should have the FASTQ file accession associated with the replicate. This can be found in the RNA-Seq Data Matrix. If multiple data files are combined in an expression call, the column name is a comma-separated list of the accessions.
• Quantification values reported: Units must be TPM, NA values are allowed when no expression is observed.

Gene expression will be calculated by summing up all the transcripts’ expression values coming from the same locus.

Expression matrix example

ID	ENCFF155CFF	ENCFF771DIX	ENCFF600LIU
PB.1.1	49.98	49.22	75.64
PB.10.1	39.67   42.66   4.74
PB.10.2	33.08   29.32   5.37
PB.10.3	0.38    0.77    1.54
PB.100.1        37.57   36.33   18.97
PB.1000.1       184.49  157.44  44.97
PB.1000.3       1.46    1.46    0.43
PB.1000.4       7.21    7.50    4.95
PB.1000.5       13.38   14.55   0.33
PB.1000.6       2.53    6.68    8.29
PB.1000.7       3.44    2.14    2.31
PB.1000.9       48.80   66.51   31.94
PB.1001.1       1044.55 1134.42 382.44
PB.1002.1       342.16  348.46  148.90
PB.1002.2       882.45  926.07  575.31
PB.1003.1       19.04   21.93   30.88
PB.1003.2       22.54   15.69   76.57

This site is open source. Improve this page.