NMD variant effect prediction

The NMD-Scanner is a Python-based variant effect annotation tool that predicts the likelihood of transcript degradation through nonsense-mediated decay (NMD). It reconstructs reference and alternative coding sequences as well as transcript sequences in some cases, identifies premature termination codons (PTCs), and evaluates canonical and non-canonical NMD escape rules. It can handle single-nucleotide variants, multiple base substitutions, long and short deletions and duplications as well as frameshift variants.

Features

• Reconstructs reference and alternative CDS, reference transcript sequence and (in some cases) the alternative transcript sequences with metadata
• Detects start / stop-loss and premature termination codons (PTCs) with the exact position in the CDS and in which exon it lies
• Computes different NMD-related features:
  • Total, upstream and downstream exon count
  • Distance of PTC to original stop codon
  • Distance of PTC to start codon
  • Transcript length
  • 3' and 5' UTR lengths
• Evaluates five canonical NMD escape rules:
  • Last exon rule
  • 50nt penultimate rule
  • Long exon rule
  • Start-proximal rule
  • Single-exon rule
• Outputs all annotations as a structured DataFrame (CSV)

Installation

Requires Python >= 3.12.

From PyPI:

pip install nmd-scanner

Writing Parquet output additionally requires pyarrow (pip install pyarrow); it is not pulled in by default.

Usage

Option 1: Annotating a VCF on the command line

After pip install . the nmd-scanner command is available:

nmd-scanner --vcf input.vcf --gtf annotation.gtf --fasta reference.fa --output results/input.csv

# write Parquet instead of CSV (requires pyarrow)
nmd-scanner --vcf input.vcf --gtf annotation.gtf --fasta reference.fa --output results/input.parquet

# option: fix exon numbering (recommended for hg19)
nmd-scanner --vcf input.vcf --gtf annotation.gtf --fasta reference.fa --output results/input.csv --reassign_exons

The equivalent python -m nmd_scanner.cli ... invocation also works without installing the console script.

Arguments:

• --vcf: Path to input VCF (SNVs / Indels supported; frameshifts handled)
• --gtf: Path to gene annotation (GTF)
• --fasta: Path to reference genome FASTA
• --output: Path to the output file. Extension selects the format: .csv for CSV, .parquet or .pq for Parquet. The parent directory must already exist; the file is overwritten if present.
• --reassign_exons: (flag) Recompute exon numbers (useful for hg19)

Output:

• The file specified by --output, containing:
  • reconstructed reference / alternative CDS and transcript sequences (+ metadata)
  • PTC detection and start / stop-loss flags
  • NMD escape rules
  • extra features such as UTR lengths, exon counts, distances, etc.

Option 2: Import as a python module

Instead of running the entire pipeline, you can import NMD-Scanner in Python and call only specific components. This is useful if you want to

• only reconstruct transcript / CDS sequences
• only compute NMD escape rules
• integrate NMD-Scanner into a larger workflow
• build custom features

For reconstructing reference and alternative coding and transcript sequences, PTC detection and start / stop-loss information:

import pandas as pd
from pyfaidx import Fasta

import nmd_scanner

vcf = nmd_scanner.read_vcf("input.vcf")
gtf_pr = nmd_scanner.read_gtf("annotation.gtf")
fasta = Fasta("reference.fa")

# Optional: fix exon numbering (recommended for hg19)
gtf_pr = nmd_scanner.compute_exon_numbers(gtf_pr)

gtf_df = gtf_pr.df
cds_df = gtf_df[gtf_df["Feature"] == "CDS"]
exons_df = gtf_df[gtf_df["Feature"] == "exon"].copy()
exons_df["exon_length"] = exons_df["End"] - exons_df["Start"]

results = nmd_scanner.extract_ptc(cds_df, vcf, fasta, exons_df)

Add extra NMD-related features (utr lengths, exon counts, ptc-related features) to the above computed results. Run this before the escape rules: evaluate_nmd_escape_rules reads the exon-count and ptc-exon-length columns produced here.

extra_features = results.apply(nmd_scanner.add_nmd_features, axis=1, result_type='expand')
results = pd.concat([results, extra_features], axis=1)

Add NMD escape rules (last exon rule, 50 nt penultimate rule, long exon rule, start proximal rule, single exon rule, nmd escape) to the above computed results:

nmd_results = results.apply(nmd_scanner.evaluate_nmd_escape_rules, axis=1, result_type='expand')
results = pd.concat([results, nmd_results], axis=1)

License

All source code in this repository is licensed under the MIT License.

Citation

Schröder, C.H. (2025). Enhanced Aberrant Gene Expression Prediction across Human Tissues. Master's Thesis, Technical University of Munich / Ludwig-Maximilians-Universität München.