C1orf112: Chromosome 1 Open Reading Frame 112
ENSG00000000460 (Ensembl) - Q9NSG2 (CA112_HUMAN) (UniProt)
Overview
This case study demonstrates the QSplice module of TRIFID, which quantifies splice junction coverage from RNA-seq data. C1orf112 serves as an excellent example of how RNA-seq evidence contributes to isoform functionality predictions.
What is QSplice?
QSplice is a TRIFID module that:
- Quantifies splice junction coverage from STAR RNA-seq alignments
- Maps unique reads to genome positions using collapsed coding splice junctions
- Calculates coverage scores per transcript
- Integrates with TRIFID’s machine learning model as predictive features
QSplice Methodology
- Genome annotation: GENCODE GFF3 file
- RNA-seq samples: STAR
SJ.out.tab files from E-MTAB-2836
- 32 human tissues
- 122 individuals
- Comprehensive tissue expression atlas
Running QSplice
python -m trifid.preprocessing.qsplice \
--gff data/external/genome_annotation/GRCh38/g27/gencode.v27.annotation.gff3.gz \
--outdir data/external/qsplice/GRCh38/g27 \
--samples out/E-MTAB-2836/GRCh38/STAR/g27 \
--version g
Output Files
sj_maxp.emtab2836.mapped.tsv.gz: Splice junction-level scoresqsplice.emtab2836.g27.tsv.gz: Transcript-level scores (TRIFID input)
C1orf112 Splice Junction Analysis
Splice Junction Scores for ENST00000472795
| Chromosome | Type | Start | End | Strand | Gene ID | Gene Name | Transcript ID | CDS Coverage | Intron # | Unique Reads | Tissue | Gene Mean | Gene Mean CDS | RNA2sj | RNA2sj_cds |
|---|
| chr1 | intron | 169794906 | 169798856 | + | ENSG00000000460 | C1orf112 | ENST00000472795 | none | 1 | 2 | tonsil | 67.37 | 73.78 | 0.0297 | 0.0271 |
| chr1 | intron | 169798959 | 169800882 | + | ENSG00000000460 | C1orf112 | ENST00000472795 | none | 2 | 69 | testis | 67.37 | 73.78 | 1.024 | 0.9352 |
| chr1 | intron | 169800972 | 169802620 | + | ENSG00000000460 | C1orf112 | ENST00000472795 | full | 3 | 74 | testis | 67.37 | 73.78 | 1.098 | 1.0029 |
| chr1 | intron | 169802726 | 169803168 | + | ENSG00000000460 | C1orf112 | ENST00000472795 | full | 4 | 77 | testis | 67.37 | 73.78 | 1.143 | 1.0436 |
| chr1 | intron | 169803310 | 169804074 | + | ENSG00000000460 | C1orf112 | ENST00000472795 | full | 5 | 57 | testis | 67.37 | 73.78 | 0.846 | 0.7725 |
Key Observations
- Maximum coverage selection: QSplice selects the junction with maximum coverage (intron 5) across all tissues
- Unique reads: Junction 5 has 57 unique reads in testis (the tissue with highest expression)
- Minimum bottleneck: This junction represents the lowest coverage among coding splice junctions for this isoform
- Normalized score: RNA2sj = 57 / 67.37 = 0.846
QSplice identifies the “weakest link” in the splice junction chain, providing a conservative estimate of transcript expression.
Transcript-Level QSplice Scores
| Chromosome | Gene ID | Gene Name | Transcript ID | Intron # | Exons | CDS Exons | Unique Reads | Tissue | RNA2sj | RNA2sj_cds |
|---|
| chr1 | ENSG00000000460 | C1orf112 | ENST00000286031 | 6 | 24 | 22 | 53 | testis | 0.787 | 0.718 |
| chr1 | ENSG00000000460 | C1orf112 | ENST00000359326 | 7 | 25 | 22 | 53 | testis | 0.787 | 0.718 |
| chr1 | ENSG00000000460 | C1orf112 | ENST00000413811 | 20 | 23 | 14 | 62 | testis | 0.920 | 0.840 |
| chr1 | ENSG00000000460 | C1orf112 | ENST00000459772 | 2 | 23 | 3 | 7 | fallopian tube | 0.104 | 0.095 |
| chr1 | ENSG00000000460 | C1orf112 | ENST00000466580 | 2 | 8 | 3 | 7 | fallopian tube | 0.104 | 0.095 |
| chr1 | ENSG00000000460 | C1orf112 | ENST00000472795 | 5 | 6 | 4 | 57 | testis | 0.846 | 0.773 |
| chr1 | ENSG00000000460 | C1orf112 | ENST00000481744 | 2 | 7 | 3 | 7 | fallopian tube | 0.104 | 0.095 |
| chr1 | ENSG00000000460 | C1orf112 | ENST00000496973 | 5 | 6 | 6 | 8 | tonsil | 0.119 | 0.108 |
| chr1 | ENSG00000000460 | C1orf112 | ENST00000498289 | 3 | 29 | 0 | 0 | - | 0 | 0 |
Interpretation
- ENST00000472795: Moderate RNA2sj score (0.846) indicates good expression support
- Tissue specificity: Highest expression in testis
- Low-scoring isoforms: Some isoforms show minimal expression (< 0.1), suggesting limited functional relevance
- Non-coding isoform: ENST00000498289 has 0 CDS exons and no expression
Visual Representation
ENST00000472795 Exon Structure
Exon 1 |Intron 1| Exon 2 |Intron 2| Exon 3 |Intron 3| Exon 4 |Intron 4| Exon 5 |Intron 5| Exon 6
| 2 | 69 | 74 | 77 | 57* |
5'UTR CDS Start Bottleneck CDS End
QSplice Features in TRIFID
QSplice generates two main features used in TRIFID:
- RNA2sj: Unique reads divided by gene average (all splice junctions)
- RNA2sj_cds: Unique reads divided by gene average (only CDS-spanning junctions)
These features contribute to TRIFID predictions by providing:
- Expression evidence for isoform existence
- Tissue-specific functional context
- Quantitative support beyond annotation
Running QSplice on Your Data
Using STAR SJ.out.tab Files
# With pre-computed STAR alignments
python -m trifid.preprocessing.qsplice \
--gff your_annotation.gff3.gz \
--outdir output/qsplice \
--samples path/to/star/output \
--version g
Using Custom Splice Junction File
# With custom SJ.out.tab
python -m trifid.preprocessing.qsplice \
--gff your_annotation.gff3.gz \
--outdir output/qsplice \
--custom custom_SJ.out.tab \
--version g
Integration with TRIFID Predictions
QSplice scores are integrated into the full TRIFID feature set alongside:
- APPRIS structural annotations
- PhyloCSF conservation scores
- Pfam domain effects
- Transcript Support Levels (TSL)
- GENCODE basic annotation
The Random Forest model learns the importance of RNA-seq evidence relative to other features, automatically weighting expression support appropriately.
Pre-computed QSplice Data
Pre-computed QSplice scores are available for:
- GENCODE 27 (Human, GRCh38)
- GENCODE 42 (Human, GRCh38)
- GENCODE 25 (Mouse, GRCm38)
- Various other genome versions
See the Data Availability page for download links.
Code Example: Analyzing QSplice Results
import pandas as pd
import matplotlib.pyplot as plt
# Load QSplice results
qsplice = pd.read_csv(
'data/external/qsplice/GRCh38/g27/qsplice.emtab2836.g27.tsv.gz',
sep='\t',
compression='gzip'
)
# Filter for gene of interest
gene_data = qsplice[qsplice['gene_name'] == 'C1orf112']
# Plot RNA2sj scores
fig, ax = plt.subplots(figsize=(10, 6))
ax.barh(gene_data['transcript_id'], gene_data['RNA2sj'])
ax.set_xlabel('RNA2sj Score')
ax.set_ylabel('Transcript ID')
ax.set_title('C1orf112 Splice Junction Coverage Scores')
ax.axvline(x=0.5, color='r', linestyle='--', label='Functional threshold')
ax.legend()
plt.tight_layout()
plt.show()
References
Next Steps
