Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Bio Genome Intervals Coverage Analysis · FreedomIntelligence | DecimalAI

Bio Genome Intervals Coverage Analysis

SkillScore94Maintained

bio-genome-intervals-coverage-analysis

Calculate read depth and coverage across genomic intervals using bedtools genomecov and coverage. Generate bedGraph files, compute per-base depth, and summarize coverage statistics. Use when assessing sequencing depth, creating coverage tracks, or evaluating target capture efficiency.

scienceby FreedomIntelligence

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

Data Analyst pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/bio-genome-intervals-coverage-analysis/SKILL.md

decimalai skills pull bio-genome-intervals-coverage-analysis --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/bio-genome-intervals-coverage-analysis/SKILL.md https://app.decimal.ai/s/bio-genome-intervals-coverage-analysis/SKILL.md

Raw file: /s/bio-genome-intervals-coverage-analysis/SKILL.md ↗ · machine-readable summary at /s/bio-genome-intervals-coverage-analysis.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

Runs

233

all versions · 30d

Eval pass

100%

global

Used when offered

100%

used / offered · all versions

By model· real traffic, every org · 30d — not the benchmark

Model	Eval pass	Runs
gemini-3.6-flash	100%	12

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

22 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 49d ago

Yes — +5% lift · at higher cost+140% tokens0% turns91% of whole cases pass with skill (20/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-09	✗→✓	▲ Improved	77%	0%
case-11	✓→✓	＝ Same ✓	75%	0%
case-06	✓→✓	＝ Same ✓	342%	0%
case-16	✓→✓	＝ Same ✓	130%	0%
case-17	✓→✓	＝ Same ✓	226%	0%

Showing 5 of 22 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/1/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

FreedomIntelligence / OpenClaw-Medical-Skills

skills/bio-genome-intervals-coverage-analysis/SKILL.md

View on GitHub ↗

Imported 8/1/2026Version v2

Skill contents

SKILL.md

<!--

COPYRIGHT NOTICE

This file is part of the "Universal Biomedical Skills" project.

Copyright (c) 2026 MD BABU MIA, PhD <md.babu.mia@mssm.edu>

All Rights Reserved.

This code is proprietary and confidential.

Unauthorized copying of this file, via any medium is strictly prohibited.

Provenance: Authenticated by MD BABU MIA

-->

Coverage Analysis

Calculate coverage and depth across genomic regions using bedtools and pybedtools.

genomecov - Genome-wide Coverage

Per-base Coverage (bedGraph)

bash
# Generate bedGraph from BAM (per-base depth)
bedtools genomecov -ibam alignments.bam -bg > coverage.bedGraph

# Include zero-coverage regions
bedtools genomecov -ibam alignments.bam -bga > coverage_with_zeros.bedGraph

# Split by strand
bedtools genomecov -ibam alignments.bam -bg -strand + > plus_strand.bedGraph
bedtools genomecov -ibam alignments.bam -bg -strand - > minus_strand.bedGraph

# Scale by total reads (RPM normalization)
TOTAL=$(samtools view -c alignments.bam)
SCALE=$(echo "scale=10; 1000000/$TOTAL" | bc)
bedtools genomecov -ibam alignments.bam -bg -scale $SCALE > normalized.bedGraph

# Use only 5' end of reads
bedtools genomecov -ibam alignments.bam -bg -5 > five_prime.bedGraph

# Use only 3' end of reads
bedtools genomecov -ibam alignments.bam -bg -3 > three_prime.bedGraph

Coverage Histogram

bash
# Genome-wide coverage histogram
bedtools genomecov -ibam alignments.bam > coverage_hist.txt

# Output format: chr, depth, bases_at_depth, chr_size, fraction
# genome  0       1000000  10000000  0.1
# genome  1       5000000  10000000  0.5
# ...

Coverage from BED

bash
# Coverage from BED intervals
bedtools genomecov -i regions.bed -g genome.txt -bg > coverage.bedGraph

# BED must be sorted
bedtools sort -i regions.bed | bedtools genomecov -i stdin -g genome.txt -bg > coverage.bedGraph

Python

python
import pybedtools

# From BAM
bam = pybedtools.BedTool('alignments.bam')
coverage = bam.genome_coverage(bg=True)
coverage.saveas('coverage.bedGraph')

# With zeros
coverage = bam.genome_coverage(bga=True)

# Normalized
coverage = bam.genome_coverage(bg=True, scale=0.001)

# From BED
bed = pybedtools.BedTool('regions.bed')
coverage = bed.genome_coverage(bg=True, g='genome.txt')

coverage - Coverage per Feature

Basic Coverage

bash
# Calculate how much of each region in A is covered by B
bedtools coverage -a targets.bed -b reads.bed > coverage_per_target.bed

# Output adds 4 columns: overlaps, bases_covered, region_length, fraction
# chr1  100  200  region1  5  50  100  0.5

# From BAM
bedtools coverage -a targets.bed -b alignments.bam > coverage.bed

# Count only (no coverage calculation)
bedtools coverage -a targets.bed -b reads.bed -counts > counts.bed

Coverage Options

bash
# Mean coverage per region
bedtools coverage -a targets.bed -b alignments.bam -mean > mean_coverage.bed

# Same strand only
bedtools coverage -a targets.bed -b alignments.bam -s > same_strand.bed

# Report depth at each position (histogram)
bedtools coverage -a targets.bed -b alignments.bam -d > per_base.bed

# Require minimum overlap
bedtools coverage -a targets.bed -b reads.bed -f 0.5 > min_overlap.bed

# Split alignments (for RNA-seq)
bedtools coverage -a exons.bed -b alignments.bam -split > exon_coverage.bed

Python

python
import pybedtools

a = pybedtools.BedTool('targets.bed')
b = pybedtools.BedTool('alignments.bam')

# Basic coverage
result = a.coverage(b)

# Mean coverage
result = a.coverage(b, mean=True)

# Counts only
result = a.coverage(b, counts=True)

# Per-base depth
result = a.coverage(b, d=True)

result.saveas('coverage.bed')

multicov - Counts Across Multiple BAMs

bash
# Count reads in regions across multiple samples
bedtools multicov -bams sample1.bam sample2.bam sample3.bam -bed regions.bed > counts.txt

# Require mapping quality
bedtools multicov -bams sample1.bam sample2.bam -bed regions.bed -q 30 > counts.txt

# Split alignments
bedtools multicov -bams sample1.bam sample2.bam -bed regions.bed -s -split > counts.txt

Calculate Coverage Statistics

Mean/Median Depth

python
import pybedtools
import pandas as pd
import numpy as np

# Load coverage BED (from bedtools coverage -d)
bed = pybedtools.BedTool('per_base_coverage.bed')
df = bed.to_dataframe()

# Calculate stats per region
stats = df.groupby(['chrom', 'start', 'end']).agg({
    'score': ['mean', 'median', 'std', 'max']
}).reset_index()

print(stats)

Coverage Distribution

python
import pybedtools

# Get coverage histogram
bam = pybedtools.BedTool('alignments.bam')
hist = bam.genome_coverage()

# Parse histogram
depths = []
fractions = []
for line in open(hist.fn):
    fields = line.strip().split('\t')
    if fields[0] == 'genome':
        depths.append(int(fields[1]))
        fractions.append(float(fields[4]))

# Calculate metrics
import numpy as np
mean_depth = sum(d * f for d, f in zip(depths, fractions))
print(f'Mean depth: {mean_depth:.1f}x')

Common Patterns

Target Region Coverage Summary

bash
# Get per-region coverage stats
bedtools coverage -a targets.bed -b alignments.bam | \
    awk -v OFS='\t' '{
        mean = ($NF > 0) ? $5/$6 : 0;
        print $1, $2, $3, $4, $7, mean
    }' > summary.bed

# Regions with low coverage
bedtools coverage -a targets.bed -b alignments.bam | \
    awk '$NF < 0.8' > low_coverage.bed

Normalize to CPM (Counts Per Million)

python
import pybedtools

bam = pybedtools.BedTool('alignments.bam')

# Get total reads
import subprocess
result = subprocess.run(['samtools', 'view', '-c', 'alignments.bam'],
                        capture_output=True, text=True)
total_reads = int(result.stdout.strip())

# Generate CPM-normalized bedGraph
scale_factor = 1000000 / total_reads
coverage = bam.genome_coverage(bg=True, scale=scale_factor)
coverage.saveas('cpm_normalized.bedGraph')

Exon Coverage for RNA-seq

bash
# Calculate coverage across exons (handling spliced reads)
bedtools coverage -a exons.bed -b alignments.bam -split > exon_coverage.bed

# Summarize by gene
awk -v OFS='\t' '{
    gene = $4; gsub(/_exon.*/, "", gene);
    sum[gene] += $NF * ($3-$2);
    len[gene] += $3-$2;
}
END {
    for (g in sum) print g, sum[g]/len[g];
}' exon_coverage.bed > gene_coverage.txt

bedGraph Format

# bedGraph: chr, start, end, value (0-based coordinates)
chr1    0       100     0
chr1    100     200     5.5
chr1    200     300     10.2
chr1    300     400     3.1

Key Parameters

| Tool | Parameter | Description | |------|-----------|-------------| | genomecov -bg | bedGraph | Output bedGraph format | | genomecov -bga | bedGraph all | Include zero coverage | | genomecov -scale | Normalize | Scale values by factor | | coverage -mean | Mean | Report mean coverage | | coverage -d | Per-base | Report per-position depth | | coverage -counts | Count | Count overlaps only | | multicov -q | Quality | Minimum mapping quality |

Related Skills

bigwig-tracks - Convert bedGraph to bigWig
alignment-files/sam-bam-basics - BAM processing
interval-arithmetic - Intersect with regions
chip-seq/chipseq-visualization - Peak coverage analysis

Illustrative Want to see what bio-genome-intervals-coverage-analysis would put in front of your model? Preview the rendered body and check its trigger phrases — no LLM call, no signup.

Latest benchmark · publisher runVerified

8/2/2026, 16:23 UTC · 22 cases

Lift

+5%

91% vs 86%

Agent turns

1.0 vs 1.0

Tokens

+140%

2,958.82 vs 1,234.32

fail → pass · 1pass → pass · 19fail → fail · 2

Per-case results (22)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-11 You are running coverage on targets.bed and alignments.bam. You only want to output the number of overlapping reads per target feature without calculating base-level coverage or fractional coverage. Which option switch provides this?	pass→pass	8,230	1,683	-80%	1	1	0%	1,431	2,505	+75%	0	0	—
case-06 You have a BAM file with 20,000,000 mapped reads. You want to generate an RPM-normalized bedGraph file using genomecov. What numerical scaling factor must be calculated and passed to the command?	pass→pass	3,290	3,255	-1%	1	1	0%	655	2,896	+342%	0	0	—
case-16 Write a command to simultaneously quantify read counts across three BAM files (s1.bam, s2.bam, s3.bam) for regions in targets.bed, filtering out reads with a mapping quality score under 30.	pass→pass	6,302	2,927	-54%	1	1	0%	1,235	2,844	+130%	0	0	—
case-17 Using the Python pybedtools package, given BedTool objects `a` (targets) and `b` (alignments), write the Python code method call to calculate mean coverage depth for each target region.	pass→pass	4,564	2,891	-37%	1	1	0%	824	2,686	+226%	0	0	—
case-18 Write a Python snippet using pybedtools to generate a CPM-normalized bedGraph from a BAM object `bam` when the calculated scaling factor is stored in variable `sf`.	pass→pass	4,677	3,063	-35%	1	1	0%	863	2,887	+235%	0	0	—
case-01 I have a genome coverage bedGraph file named depth.bedGraph and a chromosome sizes file named hg38.chrom.sizes. Provide the CLI command to convert this bedGraph file into an indexed binary bigWig track file named depth.bw.	pass→pass	3,323	2,696	-19%	1	1	0%	644	2,770	+330%	0	0	—
case-02 I have a ChIP-seq treatment BAM file treat.bam and control BAM file control.bam aligned to human genome hg38. What command line tool and subcommand should be called to identify narrow enrichment peaks?	fail→fail	7,240	3,518	-51%	1	1	0%	1,347	2,864	+113%	0	0	—
case-03 I need to generate genotype likelihoods in BCF format across target genomic regions using an aligned sample BAM file sample.bam and reference fasta ref.fa. What command line tool handles this pileup generation?	pass→pass	7,966	5,235	-34%	1	1	0%	1,617	3,172	+96%	0	0	—
case-04 Write a command line call using genomecov to generate a bedGraph file from alignments.bam that explicitly includes genomic intervals with zero depth of coverage.	pass→pass	3,057	2,204	-28%	1	1	0%	546	2,627	+381%	0	0	—
case-05 When processing a BAM file with genomecov, how do you restrict the per-base bedGraph coverage output strictly to reads originating from the forward strand?	pass→pass	10,300	2,881	-72%	1	1	0%	1,761	2,813	+60%	0	0	—
case-07 To analyze 5' end read density in a genomic alignment file alignments.bam, what flag should be supplied to genomecov so only the 5' terminal coordinate of each read contributes to coverage?	pass→pass	2,492	2,403	-4%	1	1	0%	371	2,606	+602%	0	0	—
case-08 When genomecov outputs a genome-wide depth histogram without bedGraph options, each summary row for genome contains depth in column 2 and fraction of genome at that depth in column 5. Describe how to calculate the overall mean genome depth from these histogram values.	pass→pass	8,979	7,282	-19%	1	1	0%	1,703	3,753	+120%	0	0	—
case-09 You want to compute genomic coverage from a BED file of regions named features.bed and genome sizes file genome.txt using genomecov. If features.bed is currently unsorted, how must the pipeline be constructed?	fail→pass	9,616	4,100	-57%	1	1	0%	1,722	3,040	+77%	0	0	—
case-10 When running the standard coverage command with target regions BED file A and read BAM file B, four additional summary columns are appended to each target record. What specific metric is reported in the very first of these four appended columns?	pass→pass	4,502	3,971	-12%	1	1	0%	861	2,746	+219%	0	0	—
case-12 When calculating coverage of alignment BAM reads over genomic targets BED intervals, which command line option directly requests reporting the mean coverage depth across each target region?	pass→pass	4,593	1,793	-61%	1	1	0%	806	2,505	+211%	0	0	—
case-13 Which flag in the coverage tool outputs depth measurements for every individual 1-base position across all targeted regions?	pass→pass	5,832	2,005	-66%	1	1	0%	1,027	2,567	+150%	0	0	—
case-14 When computing exon coverage from RNA-seq BAM alignments that contain spliced reads with CIGAR N operations, which option ensures spliced introns are not counted as covered regions?	pass→pass	7,059	3,674	-48%	1	1	0%	1,212	2,864	+136%	0	0	—
case-15 What command parameter forces the coverage tool to only count overlaps where a minimum of 50% of the target feature is covered by the read interval?	pass→pass	7,463	2,843	-62%	1	1	0%	1,311	2,697	+106%	0	0	—
case-19 In standard bedGraph files generated by genomic interval tools, describe the zero-based or one-based coordinate system used for start and end positions.	pass→pass	7,464	5,617	-25%	1	1	0%	1,382	3,279	+137%	0	0	—
case-20 Suppose you have per-base depth output stored in a pybedtools BedTool object `bed`. How do you convert this object into a pandas DataFrame and aggregate mean and max coverage by genomic region coordinates?	pass→pass	11,014	6,240	-43%	1	1	0%	2,030	3,454	+70%	0	0	—
case-21 From standard 7-column bedtools coverage output (where col 4 is region name, col 5 is overlap count, col 6 is covered bases, col 7 is region length), write an awk expression to calculate the mean depth per region.	fail→fail	16,521	13,952	-16%	1	1	0%	2,993	5,002	+67%	0	0	—
case-22 What dedicated command line utility allows passing multiple BAM alignment files simultaneously using `-bams` to generate a single region-by-sample count matrix against a target BED file?	pass→pass	4,395	1,575	-64%	1	1	0%	814	2,517	+209%	0	0	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +5 percentage points is the difference between those two pass rates over the 22 comparable cases.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 86.4%
Pass rate with the skill: 90.9%
Cases attempted: 22
Cases counted toward lift: 22
Cases that regressed: 0
Skill version measured: v2
Run completed: 8/2/2026, 16:23 UTC

Measured example — same prompt, with vs. without the skill · from the benchmark run · case case-09

Prompt

You want to compute genomic coverage from a BED file of regions named features.bed and genome sizes file genome.txt using genomecov. If features.bed is currently unsorted, how must the pipeline be constructed?

Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.

Show both answers

SkillScore blends 4 signals3 of 4 complete

BenchmarkPass rate WITH skill on a verified run91

Live evalPass rate on real evaluated traces100%

AI ratingNeeds ≥10 rated traces

AdoptionBreadth of real use across orgs88

Efficiency · separate axis (at equal correctness)+0% turns+157% tokens

Shown for transparency; not credited unless a verified run shows no correctness regression.

Backed by 3 of 4 signals — each added signal makes the score more robust.

Verified on gemini-3.6-flash · re-tested 48 days ago

Model	Method	Date	Lift
gemini-3.6-flash	verified	7/26/2026	+5%

Other measured skills in the registry, with their headline benchmark lift.

Ready to supercharge your AI agents?

Bio Genome Intervals Coverage Analysis

SkillSafety

Source

COPYRIGHT NOTICE

This file is part of the "Universal Biomedical Skills" project.

Copyright (c) 2026 MD BABU MIA, PhD <md.babu.mia@mssm.edu>

All Rights Reserved.

This code is proprietary and confidential.

Unauthorized copying of this file, via any medium is strictly prohibited.

Provenance: Authenticated by MD BABU MIA

Coverage Analysis

genomecov - Genome-wide Coverage

Per-base Coverage (bedGraph)

Coverage Histogram

Coverage from BED

Python

coverage - Coverage per Feature

Basic Coverage

Coverage Options

Python

multicov - Counts Across Multiple BAMs

Calculate Coverage Statistics

Mean/Median Depth

Coverage Distribution

Common Patterns

Target Region Coverage Summary

Normalize to CPM (Counts Per Million)

Exon Coverage for RNA-seq

bedGraph Format

Key Parameters

Related Skills

How this skill was measured

Measured example — same prompt, with vs. without the skill · from the benchmark run · case case-09

Related skills