Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Align RNA-seq reads with STAR (Spliced Transcripts Alignment to a Reference). Supports two-pass mode for novel splice junction discovery. Use when aligning RNA-seq data requiring splice-aware alignment.
.claude/skills/bio-read-alignment-star-alignment/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 48% | 0% |
| case-09 | ✓→✓ | ＝ Same ✓ | 282% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | 105% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | 73% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 284% | 0% |
<!--
#
#
-->
bash# Basic index generation STAR --runMode genomeGenerate \ --runThreadN 8 \ --genomeDir star_index/ \ --genomeFastaFiles reference.fa \ --sjdbGTFfile annotation.gtf \ --sjdbOverhang 100 # Read length - 1
bash# For 150bp reads, use sjdbOverhang=149 STAR --runMode genomeGenerate \ --runThreadN 8 \ --genomeDir star_index_150/ \ --genomeFastaFiles reference.fa \ --sjdbGTFfile annotation.gtf \ --sjdbOverhang 149
bash# Paired-end alignment STAR --runThreadN 8 \ --genomeDir star_index/ \ --readFilesIn reads_1.fq.gz reads_2.fq.gz \ --readFilesCommand zcat \ --outFileNamePrefix sample_ \ --outSAMtype BAM SortedByCoordinate
bashSTAR --runThreadN 8 \ --genomeDir star_index/ \ --readFilesIn reads.fq.gz \ --readFilesCommand zcat \ --outFileNamePrefix sample_ \ --outSAMtype BAM SortedByCoordinate
bash# Two-pass mode for better novel junction detection STAR --runThreadN 8 \ --genomeDir star_index/ \ --readFilesIn r1.fq.gz r2.fq.gz \ --readFilesCommand zcat \ --outFileNamePrefix sample_ \ --outSAMtype BAM SortedByCoordinate \ --twopassMode Basic
bash# Output gene counts (like featureCounts) STAR --runThreadN 8 \ --genomeDir star_index/ \ --readFilesIn r1.fq.gz r2.fq.gz \ --readFilesCommand zcat \ --outFileNamePrefix sample_ \ --outSAMtype BAM SortedByCoordinate \ --quantMode GeneCounts
Output: sample_ReadsPerGene.out.tab with columns:
bash# ENCODE recommended settings STAR --runThreadN 8 \ --genomeDir star_index/ \ --readFilesIn r1.fq.gz r2.fq.gz \ --readFilesCommand zcat \ --outFileNamePrefix sample_ \ --outSAMtype BAM SortedByCoordinate \ --outSAMunmapped Within \ --outSAMattributes NH HI AS NM MD \ --outFilterType BySJout \ --outFilterMultimapNmax 20 \ --outFilterMismatchNmax 999 \ --outFilterMismatchNoverReadLmax 0.04 \ --alignIntronMin 20 \ --alignIntronMax 1000000 \ --alignMatesGapMax 1000000 \ --alignSJoverhangMin 8 \ --alignSJDBoverhangMin 1
bash# For chimeric/fusion detection STAR --runThreadN 8 \ --genomeDir star_index/ \ --readFilesIn r1.fq.gz r2.fq.gz \ --readFilesCommand zcat \ --outFileNamePrefix sample_ \ --outSAMtype BAM SortedByCoordinate \ --chimSegmentMin 12 \ --chimJunctionOverhangMin 8 \ --chimOutType Junctions WithinBAM SoftClip \ --chimMainSegmentMultNmax 1
| File | Description | |------|-------------| | Aligned.sortedByCoord.out.bam | Sorted BAM file | | Log.final.out | Alignment summary statistics | | Log.out | Detailed log | | SJ.out.tab | Splice junctions | | ReadsPerGene.out.tab | Gene counts (if --quantMode) | | Chimeric.out.junction | Fusion candidates (if chimeric) |
bash# Reduce memory for limited systems STAR --genomeLoad NoSharedMemory \ --limitBAMsortRAM 10000000000 \ # 10GB for sorting ... # For very large genomes, limit during index generation STAR --runMode genomeGenerate \ --limitGenomeGenerateRAM 31000000000 \ # 31GB ...
bash# Load genome into shared memory (for multiple samples) STAR --genomeLoad LoadAndExit --genomeDir star_index/ # Run alignments (faster startup) STAR --genomeLoad LoadAndKeep --genomeDir star_index/ ... # Remove from memory when done STAR --genomeLoad Remove --genomeDir star_index/
| Parameter | Default | Description | |-----------|---------|-------------| | --runThreadN | 1 | Number of threads | | --sjdbOverhang | 100 | Read length - 1 | | --outFilterMultimapNmax | 10 | Max multi-mapping | | --alignIntronMax | 0 | Max intron size | | --outFilterMismatchNmax | 10 | Max mismatches | | --outSAMtype | SAM | Output format | | --quantMode | - | GeneCounts for counting | | --twopassMode | None | Basic for two-pass |
<!-- AUTHOR_SIGNATURE: 9a7f3c2e-MD-BABU-MIA-2026-MSSM-SECURE -->
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-22 | fail→fail | 8,749 | 9,077 | +4% | 1 | 1 | 0% | 1,855 | 3,378 | +82% | 0 | 0 | — |
case-09 | pass→pass | 2,903 | 2,377 | -18% | 1 | 1 | 0% | 521 | 1,990 | +282% | 0 | 0 | — |
case-01 | fail→pass | 8,915 | 5,117 | -43% | 1 | 1 | 0% | 1,777 | 2,632 | +48% | 0 | 0 | — |
case-02 | pass→pass | 6,597 | 4,929 | -25% | 1 | 1 | 0% | 1,205 | 2,469 | +105% | 0 | 0 | — |
case-03 | pass→pass | 6,729 | 3,262 | -52% | 1 | 1 | 0% | 1,266 | 2,187 | +73% | 0 | 0 | — |
case-04 | pass→pass | 2,769 | 2,927 | +6% | 1 | 1 | 0% | 547 | 2,100 | +284% | 0 | 0 | — |
case-05 | pass→pass | 4,807 | 2,392 | -50% | 1 | 1 | 0% | 812 | 1,989 | +145% | 0 | 0 | — |
case-06 | pass→pass | 3,812 | 3,190 | -16% | 1 | 1 | 0% | 708 | 2,218 | +213% | 0 | 0 | — |
case-07 | pass→pass | 4,372 | 6,903 | +58% | 1 | 1 | 0% | 801 | 2,261 | +182% | 0 | 0 | — |
case-08 | pass→pass | 2,741 | 2,007 | -27% | 1 | 1 | 0% | 484 | 1,882 | +289% | 0 | 0 | — |
case-10 | pass→pass | 9,625 | 3,353 | -65% | 1 | 1 | 0% | 1,723 | 2,130 | +24% | 0 | 0 | — |
case-11 | pass→pass | 4,918 | 5,935 | +21% | 1 | 1 | 0% | 884 | 2,156 | +144% | 0 | 0 | — |
case-12 | pass→pass | 3,768 | 2,149 | -43% | 1 | 1 | 0% | 686 | 1,928 | +181% | 0 | 0 | — |
case-13 | pass→pass | 9,190 | 5,562 | -39% | 1 | 1 | 0% | 1,651 | 2,490 | +51% | 0 | 0 | — |
case-14 | pass→pass | 5,013 | 2,980 | -41% | 1 | 1 | 0% | 938 | 2,072 | +121% | 0 | 0 | — |
case-15 | pass→pass | 3,135 | 2,624 | -16% | 1 | 1 | 0% | 568 | 1,920 | +238% | 0 | 0 | — |
case-16 | pass→pass | 4,071 | 3,347 | -18% | 1 | 1 | 0% | 801 | 2,183 | +173% | 0 | 0 | — |
case-17 | pass→pass | 3,536 | 2,075 | -41% | 1 | 1 | 0% | 670 | 1,926 | +187% | 0 | 0 | — |
case-18 | pass→pass | 3,237 | 2,520 | -22% | 1 | 1 | 0% | 527 | 2,065 | +292% | 0 | 0 | — |
case-19 | pass→pass | 5,006 | 3,401 | -32% | 1 | 1 | 0% | 1,010 | 2,175 | +115% | 0 | 0 | — |
case-20 | pass→pass | 6,432 | 2,658 | -59% | 1 | 1 | 0% | 1,230 | 2,011 | +63% | 0 | 0 | — |
case-21 | pass→pass | 13,588 | 12,623 | -7% | 1 | 1 | 0% | 3,004 | 4,139 | +38% | 0 | 0 | — |
case-23 | pass→pass | 5,714 | 4,118 | -28% | 1 | 1 | 0% | 1,060 | 2,349 | +122% | 0 | 0 | — |
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. 23 cases were attempted. The headline lift of +4 percentage points is the difference between those two pass rates over the 23 comparable cases.
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.
| Model | Method | Date | Lift |
|---|---|---|---|
| gemini-3.6-flash | verified | 7/26/2026 | +8% |
Other measured skills in the registry, with their headline benchmark lift.