splam 1.0.10

Splice junction scoring tool

Splam is a splice junction recognition model based on a deep residual convolutional neural network that offers fast and precise assessment of splice junctions. It was trained on combined donor-acceptor pairs and focuses on a narrow window of 400 base pairs surrounding each splice site, inspired by the understanding that the splicing process primarily depends on signals within this region.

Why Splam❓#

1. We need a tool to evaluate splice junctions & spliced alignments. Thousands of RNA-Seq datasets are generated every day, but there are no tools available for cleaning up spurious spliced alignments in these data. Splam addresses this problem!
2. Splam-cleaned alignments lead to improved transcript assembly, which, in turn, may enhance all downstream RNA-Seq analyses, including transcript quantification, differential gene expression analysis, and more.

Who is it for❓#

If you are (1) doing RNA-Seq data analysis or (2) seeking a trustworthy way to evaluate splice junctions (introns), then Splam is the tool that you are looking for!

What does Splam do❓#

There are two main use case scenarios:

1. Improving your alignment file. Splam evaluates the quality of spliced alignments and removes those containing spurious splice junctions. This significantly enhances the quality of downstream transcriptome assemblies [Link].

2. Evaluating the quality of introns in your annotation file or assembled transcripts [Link].

Documentation#

📒 The full user manual is available here

Table of contents#

• Installation
  • System requirements
  • Install through pip
  • Install from source
  • Install through conda
  • Check Splam installation
  • Now, you are ready to go !
• Quick Start Guide
  • Super-Quick Start (3 lines of code)
  • Try Splam on Google Colab
• Alignment file evaluation & cleanup (BAM)
  • Introduction
  • Workflow overview
  • Step 1: Preparing your input files
  • Step 2: Extracting splice junctions in your alignment file
  • Step 3: Scoring extracted splice junctions
  • Step 4: Cleaning up your alignment file
  • Step 5: IGV visualization
  • Step 6: Assembling alignments into transcripts
  • What's next?
• Annotation file / assembled transcripts evaluation (GFF)
  • Step 1: Preparing your input files
  • Step 2: Extracting introns in your annotation file
  • Step 3: Scoring extracted introns
  • Step 4: Evaluating isoforms by Splam scores
  • What's next?
• Splam generalizes on non-human species
  • Example: Running Splam on house mouse (Mus musculus)
  • Explanation: Splam's performance on non-human species
• Behind the scenes
  • Data curation
  • Model architecture
  • Splam training & testing
  • Reference
• Q & A ...
• User Manual
  • splam
  • splam extract
  • splam score
  • splam clean
• License
• Contact

Installation#

Splam is on PyPi. This is the easiest installation approach. Check out all the releases here.

$ pip install splam

You can also install Splam from source

$ git clone https://github.com/Kuanhao-Chao/splam --recursive

$ cd splam/src/

$ python setup.py install

Quick Start#

Running Splam is simple. It only requires three lines of code!

See these examples on Google Colab:

Example 1: clean up alignment files (BAM)

$ cd test

# Step 1: extract splice junctions in the alignment file
$ splam extract -P SRR1352129_chr9_sub.bam -o tmp_out_alignment

# Step 2: score all the extracted splice junctions
$ splam score -G chr9_subset.fa -m ../model/splam_script.pt -o tmp_out_alignment tmp_out_alignment/junction.bed

#Step 3: output a cleaned and sorted alignment file
$ splam clean -o tmp_out_alignment

Example 2: evaluate annotation files / assembled transcripts (GFF)

$ cd test

# Step 1: extract introns in the annotation
$ splam extract refseq_40_GRCh38.p14_chr_fixed.gff -o tmp_out_annotation

# Step 2: score introns in the annotation
$ splam score -G chr9_subset.fa -m ../model/splam_script.pt -o tmp_out_annotation tmp_out_annotation/junction.bed

#Step 3: output statistics of each transcript
$ splam clean -o tmp_out_annotation

Example 3: evaluate mouse annotation files (GFF)

$ cd test

# Step 1: extract introns in the annotation
$ splam extract mouse_chr19.gff -o tmp_out_generalization

# Step 2: score introns in the annotation
$ splam score -A GRCm39_assembly_report.txt -G mouse_chr19.fa -m ../model/splam_script.pt -o tmp_out_generalization tmp_out_generalization/junction.bed

# Step 3: output statistics of each transcript
$ splam clean -o tmp_out_generalization

Scripts for Splam model training & analysis#

All the scripts for Splam training and data analysis are in this GitHub repository.

Citation#

Kuan-Hao Chao*, Alan Mao, Steven L Salzberg, Mihaela Pertea*, "Splam: a deep-learning-based splice site predictor that improves spliced alignments ", bioRxiv 2023.07.27.550754, doi: https://doi.org/10.1101/2023.07.27.550754, 2023