First release.

This is the first release with the complete pipeline.

Input file:

• Probability table from NanoSplicer (TSV)
• NanoSplicer jwr_checker output (HDF5)
• BAM file (same as what has been input to NanoSplicer)

Output:

• **Option 1:**Default (or specify an output filename with extension .bed): BED file. Each entry corresponds to a unique isoform, and the "score" field (column 5) represents the read counts for each identified isoform.
• Option 2: specify an output filename with extension .gtf). Note: the quantification information is still not included in the GTF output.

Small group recover

Update:

• Start group reads based on non-overlapping interval in each chromosom
• Increase the default threshold to define groups with a sufficient number of reads
• Recover small group using all other JWRs (corrected by NanoSplicer and JAQ-based approach).

First version

This is an initial implementation of NanoIsoform:

Highlight:

1. Input data: NanoSplicer JWR checker output and prob table
   * All data are imported into memory, working okay so far but might need to adjust for larger datasets.
   * Note that read with no JWR will not appear in the input file.
2. Group reads based on "# of JWR" -> "Close JWRs"
3. No read count threshold per group, but JWR will remain uncorrected if there are not enough nearby JWR corrected by NanoSplicer or JAQ-based approach.

More details:
https://secret-meadow-572.notion.site/Software-development-c957105507df4ebda66de907a8c7212a