/bio-crispr-screens-jacks-analysis

name: bio-crispr-screens-jacks-analysis
description: JACKS (Joint Analysis of CRISPR/Cas9 Knockout Screens) for modeling sgRNA efficacy and gene essentiality. Use when analyzing multiple CRISPR screens simultaneously or when accounting for variable sgRNA efficiency across experiments.
tool_type: python
primary_tool: JACKS

Version Compatibility

Reference examples tested with: MAGeCK 0.5+, matplotlib 3.8+, numpy 1.26+, pandas 2.2+, scipy 1.12+

Before using code patterns, verify installed versions match. If versions differ:

• Python: `pip show <package>` then `help(module.function)` to check signatures
• CLI: `<tool> --version` then `<tool> --help` to confirm flags

If code throws ImportError, AttributeError, or TypeError, introspect the installed package and adapt the example to match the actual API rather than retrying.

JACKS CRISPR Screen Analysis

**"Analyze multiple CRISPR screens jointly with JACKS"** → Model sgRNA efficacy and gene essentiality simultaneously across multiple screens, accounting for variable guide efficiency.

• Python: `jacks.infer_JACKS()` for joint analysis across experiments

JACKS jointly models sgRNA efficacy and gene essentiality across multiple experiments. It infers both gene-level fitness effects and sgRNA-specific efficiency.

Installation

pip install jacks
# or
git clone https://github.com/felicityallen/JACKS.git
cd JACKS && pip install -e .

Input File Formats

Count Data

# counts.txt (tab-separated)
sgRNA	Gene	Sample1	Sample2	Sample3	Control1	Control2
sgRNA1	GENE_A	100	120	90	80	85
sgRNA2	GENE_A	200	180	210	150	160
sgRNA3	GENE_B	50	45	55	60	58
...

Replicate Map

# replicatemap.txt
Sample1	Experiment1	Day14
Sample2	Experiment1	Day14
Sample3	Experiment2	Day14
Control1	Experiment1	Day0
Control2	Experiment2	Day0

Guide-Gene Map

# guidemap.txt
sgRNA1	GENE_A
sgRNA2	GENE_A
sgRNA3	GENE_B
sgRNA4	GENE_B
...

Basic JACKS Analysis

Command Line

# Run JACKS
python -m jacks.run_JACKS \
    counts.txt \
    replicatemap.txt \
    guidemap.txt \
    output_prefix \
    --ctrl_sample_pattern "Day0" \
    --ctrl_sample_pattern_column "Condition"

Python API

**Goal:** Run JACKS joint analysis to simultaneously model sgRNA efficacy and gene essentiality across experiments.

**Approach:** Load count data, guide-gene mapping, and replicate map; separate control and treatment samples; then run MCMC inference to estimate gene fitness effects and per-sgRNA efficiency.

from jacks import infer
import pandas as pd

# Load data
counts = pd.read_csv('counts.txt', sep='\t', index_col=0)
guide_gene_map = pd.read_csv('guidemap.txt', sep='\t', header=None, names=['sgRNA', 'Gene'])
replicate_map = pd.read_csv('replicatemap.txt', sep='\t', header=None,
                             names=['Sample', 'Experiment', 'Condition'])

# Separate control and treatment samples
ctrl_samples = replicate_map[replicate_map['Condition'] == 'Day0']['Sample'].tolist()
treatment_samples = replicate_map[replicate_map['Condition'] == 'Day14']['Sample'].tolist()

# Run JACKS inference
# n_iterations=10000: MCMC iterations. Increase for final analysis.
# burn_in=1000: Burn-in period. Should be ~10% of iterations.
jacks_results = infer.run_inference(
    counts,
    guide_gene_map,
    treatment_samples,
    ctrl_samples,
    n_iterations=10000,
    burn_in=1000
)

Output Files

| File | Description | |------|-------------| | `_gene_JACKS_results.txt` | Gene-level essentiality scores | | `_grna_JACKS_results.txt` | sgRNA-level efficacy estimates | | `_jacks_full_data.pickle` | Full model for downstream analysis |

Interpret Gene Results

**Goal:** Classify genes as essential or enriched from JACKS output scores.

**Approach:** Load the gene results table, filter by JACKS score direction and FDR significance, and rank to identify top essential (negative effect) and enriched (positive effect) genes.

import pandas as pd
import numpy as np

# Load gene results
genes = pd.read_csv('output_gene_JACKS_results.txt', sep='\t')

# JACKS score: negative = essential (dropout), positive = enriched
# Columns: gene, X1 (effect), X2 (std), fdr_log10

# Essential genes (significant negative effect)
# fdr_threshold=-1: log10(FDR) < -1 means FDR < 0.1
essential = genes[(genes['X1'] < 0) & (genes['fdr_log10'] < -1)]
essential = essential.sort_values('X1')
print(f'Essential genes: {len(essential)}')
print(essential.head(20))

# Enriched genes
enriched = genes[(genes['X1'] > 0) & (genes['fdr_log10'] < -1)]
enriched = enriched.sort_values('X1', ascending=False)
print(f'Enriched genes: {len(enriched)}')

sgRNA Efficacy Analysis

**Goal:** Assess sgRNA performance to identify low-efficacy guides for library optimization.

**Approach:** Load per-sgRNA efficacy estimates from JACKS output, flag guides below an efficacy threshold, and aggregate by gene to evaluate library-level guide quality.

import pandas as pd

# Load sgRNA results
guides = pd.read_csv('output_grna_JACKS_results.txt', sep='\t')

# Efficacy scores range from 0 (ineffective) to 1 (highly effective)
# X1 column contains efficacy estimates

# Identify poor sgRNAs
# efficacy<0.3: sgRNAs with low efficacy. Consider removal in future libraries.
poor_guides = guides[guides['X1'] < 0.3]
print(f'Low efficacy guides: {len(poor_guides)}')

# Group by gene to assess library quality
gene_efficacy = guides.groupby('Gene')['X1'].agg(['mean', 'std', 'count'])
gene_efficacy = gene_efficacy.sort_values('mean')
print(gene_efficacy.head(20))

Visualization

Gene Effect Plot

import matplotlib.pyplot as plt
import numpy as np

genes = pd.read_csv('output_gene_JACKS_results.txt', sep='\t')

fig, ax = plt.subplots(figsize=(10, 8))

# Color by significance
colors = ['red' if fdr < -1 else 'gray' for fdr in genes['fdr_log10']]

ax.scatter(genes['X1'], -genes['fdr_log10'], c=colors, alpha=0.5, s=10)
ax.axhline(1, linestyle='--', co