outlier-splicing-detection

Name: outlier-splicing-detection
Author: GPTomics

1.1kstars185forks

How to install outlier-splicing-detection?

gptomics/bioskills/outlier-splicing-detection

$npx -y skills add GPTomics/bioSkills --skill outlier-splicing-detection

Installs into the current project.

Summary

Detects aberrant splicing in single rare-disease patients vs a control panel using FRASER 2.0 (Bioconductor; Beta-binomial autoencoder on Intron Jaccard Index, default delta cutoff 0.1, q hyperparameter), OUTRIDER (gene-level outlier expression via autoencoder denoising), Leafcut

Files · 1

View on GitHub

SKILL.md

1	## Version Compatibility
2
3	Reference examples tested with: FRASER 2.0 (>=1.99.0), OUTRIDER 1.20+, LeafcutterMD via leafcutter 0.2.9+, DROP 1.4+, R 4.4+, BiocManager 1.30+
4
5	Before using code patterns, verify installed versions match. If versions differ:
6	- R: `packageVersion('<pkg>')` then `?function_name` to verify parameters
7	- CLI: `<tool> --version` then `<tool> --help` to confirm flags
8
9	If code throws ImportError, AttributeError, or TypeError, introspect the installed
10	package and adapt the example to match the actual API rather than retrying.
11
12	# Outlier Splicing Detection
13
14	For clinical RNA-seq diagnostics in rare disease, the question is not "what differs between groups?" but "what is aberrant in this single patient relative to a panel of unaffected samples?". The statistical framework is single-sample-vs-cohort outlier detection, fundamentally different from two-group differential splicing. Tools in this space are designed for clinical Mendelian diagnostic settings.
15
16	## Tool Taxonomy
17
18	\| Tool \| Statistic \| Test target \| Fails when \|
19	\|------\|-----------\|-------------\|------------\|
20	\| FRASER 2.0 \| Beta-binomial autoencoder on Intron Jaccard Index \| Splicing outliers (per-sample, per-junction) \| Cohort <20 samples; tissue mismatch \|
21	\| OUTRIDER \| Autoencoder-denoised expression Z-score \| Gene-level expression outliers (LoF, monoallelic) \| Cohort <20 samples \|
22	\| LeafcutterMD \| Dirichlet-multinomial outlier mode \| Annotation-free intron usage \| Beta-binomial fits poorly OR few controls \|
23	\| DROP \| Snakemake pipeline \| All of above + monoallelic expression \| Pipeline complexity for small projects \|
24
25	Core reference: FRASER 2.0 for splicing outliers, OUTRIDER for expression outliers, DROP to combine. Standard tool in EU rare-disease programs (Solve-RD) and NIH UDN.
26
27	## Decision Tree by Diagnostic Scenario
28
29	\| Scenario \| Recommended approach \|
30	\|----------\|----------------------\|
31	\| Single rare-disease patient + panel of n>=50 controls \| FRASER 2.0 (Intron Jaccard Index) \|
32	\| Single patient + small panel (n=20-50) \| FRASER 2.0 with auxiliary GTEx controls; tune q carefully \|
33	\| Patient + cohort <20 \| Insufficient for outlier detection; consider differential or recruit more samples \|
34	\| Outlier expression suspected (loss of function, monoallelic) \| OUTRIDER on same cohort \|
35	\| Annotation-free outlier (cryptic exon, novel junction) \| LeafcutterMD \|
36	\| Integrated diagnostic pipeline (splicing + expression + MAE) \| DROP \|
37	\| TDP-43 ALS post-mortem brain (cryptic exons) \| FRASER 2.0; expect UNC13A, STMN2, ATG4B \|
38	\| SF3B1-mutant cancer sample \| FRASER 2.0 with cohort-matched RNA-seq; expect cryptic 3'ss \|
39	\| Familial dysautonomia (ELP1) \| FRASER 2.0 in fibroblast/iPSC; CNS tissue gives strongest signal \|
40	\| Stargardt deep-intronic ABCA4 \| FRASER 2.0 in retina-relevant tissue \|
41	\| Solid tumor splicing biomarker \| Differential splicing (n>=10 vs cohort) — see differential-splicing skill \|
42	\| RNA validation of SpliceAI hit \| FRASER 2.0 + cross-reference with predicted variant location \|
43
44	## When to Use Outlier vs Differential
45
46	Outlier regime (this skill):
47	- Single patient or small case series vs control panel
48	- Question: "What is aberrant in this patient?"
49	- Statistical model: single-sample p-value vs cohort distribution
50
51	Differential regime (differential-splicing skill):
52	- Two well-defined groups, n>=3 each
53	- Question: "What differs between groups?"
54	- Statistical model: two-group LRT or related
55
56	If n>=10 patients with a shared phenotype are available, prefer differential (more power); if single patient or heterogeneous case series, use outlier.
57
58	## FRASER 2.0 Workflow
59
60	Goal: Detect aberrant splicing in patient samples vs cohort using the Intron Jaccard Index.
61
62	Approach: Count split reads per junction, compute Intron Jaccard Index per intron, fit a Beta-binomial autoencoder to estimate expected values, then flag outliers by p-value and delta.
63
64	```r
65	library(FRASER); library(BiocParallel)
66
67	bam_files <- list.files('bams/', pattern='.bam$', full.names=TRUE)
68	sampl

People who liked this also viewed

6 picks

Type

outlier-splicing-detection

1.1kstars185forks

View on GitHub

How to install outlier-splicing-detection?

gptomics/bioskills/outlier-splicing-detection

$npx -y skills add GPTomics/bioSkills --skill outlier-splicing-detection

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	## Version Compatibility
2
3	Reference examples tested with: FRASER 2.0 (>=1.99.0), OUTRIDER 1.20+, LeafcutterMD via leafcutter 0.2.9+, DROP 1.4+, R 4.4+, BiocManager 1.30+
4
5	Before using code patterns, verify installed versions match. If versions differ:
6	- R: `packageVersion('<pkg>')` then `?function_name` to verify parameters
7	- CLI: `<tool> --version` then `<tool> --help` to confirm flags
8
9	If code throws ImportError, AttributeError, or TypeError, introspect the installed
10	package and adapt the example to match the actual API rather than retrying.
11
12	# Outlier Splicing Detection
13
14	For clinical RNA-seq diagnostics in rare disease, the question is not "what differs between groups?" but "what is aberrant in this single patient relative to a panel of unaffected samples?". The statistical framework is single-sample-vs-cohort outlier detection, fundamentally different from two-group differential splicing. Tools in this space are designed for clinical Mendelian diagnostic settings.
15
16	## Tool Taxonomy
17
18	\| Tool \| Statistic \| Test target \| Fails when \|
19	\|------\|-----------\|-------------\|------------\|
20	\| FRASER 2.0 \| Beta-binomial autoencoder on Intron Jaccard Index \| Splicing outliers (per-sample, per-junction) \| Cohort <20 samples; tissue mismatch \|
21	\| OUTRIDER \| Autoencoder-denoised expression Z-score \| Gene-level expression outliers (LoF, monoallelic) \| Cohort <20 samples \|
22	\| LeafcutterMD \| Dirichlet-multinomial outlier mode \| Annotation-free intron usage \| Beta-binomial fits poorly OR few controls \|
23	\| DROP \| Snakemake pipeline \| All of above + monoallelic expression \| Pipeline complexity for small projects \|
24
25	Core reference: FRASER 2.0 for splicing outliers, OUTRIDER for expression outliers, DROP to combine. Standard tool in EU rare-disease programs (Solve-RD) and NIH UDN.
26
27	## Decision Tree by Diagnostic Scenario
28
29	\| Scenario \| Recommended approach \|
30	\|----------\|----------------------\|
31	\| Single rare-disease patient + panel of n>=50 controls \| FRASER 2.0 (Intron Jaccard Index) \|
32	\| Single patient + small panel (n=20-50) \| FRASER 2.0 with auxiliary GTEx controls; tune q carefully \|
33	\| Patient + cohort <20 \| Insufficient for outlier detection; consider differential or recruit more samples \|
34	\| Outlier expression suspected (loss of function, monoallelic) \| OUTRIDER on same cohort \|
35	\| Annotation-free outlier (cryptic exon, novel junction) \| LeafcutterMD \|
36	\| Integrated diagnostic pipeline (splicing + expression + MAE) \| DROP \|
37	\| TDP-43 ALS post-mortem brain (cryptic exons) \| FRASER 2.0; expect UNC13A, STMN2, ATG4B \|
38	\| SF3B1-mutant cancer sample \| FRASER 2.0 with cohort-matched RNA-seq; expect cryptic 3'ss \|
39	\| Familial dysautonomia (ELP1) \| FRASER 2.0 in fibroblast/iPSC; CNS tissue gives strongest signal \|
40	\| Stargardt deep-intronic ABCA4 \| FRASER 2.0 in retina-relevant tissue \|
41	\| Solid tumor splicing biomarker \| Differential splicing (n>=10 vs cohort) — see differential-splicing skill \|
42	\| RNA validation of SpliceAI hit \| FRASER 2.0 + cross-reference with predicted variant location \|
43
44	## When to Use Outlier vs Differential
45
46	Outlier regime (this skill):
47	- Single patient or small case series vs control panel
48	- Question: "What is aberrant in this patient?"
49	- Statistical model: single-sample p-value vs cohort distribution
50
51	Differential regime (differential-splicing skill):
52	- Two well-defined groups, n>=3 each
53	- Question: "What differs between groups?"
54	- Statistical model: two-group LRT or related
55
56	If n>=10 patients with a shared phenotype are available, prefer differential (more power); if single patient or heterogeneous case series, use outlier.
57
58	## FRASER 2.0 Workflow
59
60	Goal: Detect aberrant splicing in patient samples vs cohort using the Intron Jaccard Index.
61
62	Approach: Count split reads per junction, compute Intron Jaccard Index per intron, fit a Beta-binomial autoencoder to estimate expected values, then flag outliers by p-value and delta.
63
64	```r
65	library(FRASER); library(BiocParallel)
66
67	bam_files <- list.files('bams/', pattern='.bam$', full.names=TRUE)
68	sampl

People who liked this also viewed

6 picks

Type