foryourhealth111-pixel/Vibe-Skills / bundled/skills/gtars
gtars
VerifiedHigh-performance toolkit for genomic interval analysis in Rust with Python bindings. Use when working with genomic regions, BED files, coverage tracks, overlap detection, tokenization for ML models, or fragment analysis in computational genomics and machine learning applications.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 7.6 KB |
| references/cli.md | markdown | 4.9 KB |
| references/coverage.md | markdown | 3.9 KB |
| references/overlap.md | markdown | 3.6 KB |
| references/python-api.md | markdown | 4.1 KB |
| references/refget.md | markdown | 3.1 KB |
| references/tokenizers.md | markdown | 2.4 KB |
references/tokenizers.md
View on GitHub# Genomic Tokenizers
Tokenizers convert genomic regions into discrete tokens for machine learning applications, particularly useful for training genomic deep learning models.
## Python API
### Creating a Tokenizer
Load tokenizer configurations from various sources:
```python
import gtars
# From BED file
tokenizer = gtars.tokenizers.TreeTokenizer.from_bed_file("regions.bed")
# From configuration file
tokenizer = gtars.tokenizers.TreeTokenizer.from_config("tokenizer_config.yaml")
# From region string
tokenizer = gtars.tokenizers.TreeTokenizer.from_region_string("chr1:1000-2000")
```
### Tokenizing Genomic Regions
Convert genomic coordinates to tokens:
```python
# Tokenize a single region
token = tokenizer.tokenize("chr1", 1000, 2000)
# Tokenize multiple regions
tokens = []
for chrom, start, end in regions:
token = tokenizer.tokenize(chrom, start, end)
tokens.append(token)
```
### Token Properties
Access token information:
```python
# Get token ID
token_id = token.id
# Get genomic coordinates
chrom = token.chromosome
start = token.start
end = token.end
# Get token metadata
metadata = token.metadata
```
## Use Cases
### Machine Learning Preprocessing
Tokenizers are essential for preparing genomic data for ML models:
1. **Sequence modeling**: Convert genomic intervals into discrete tokens for transformer models
2. **Position encoding**: Create consistent positional encodings across datasets
3. **Data augmentation**: Generate alternative tokenizations for training
### Integration with geniml
The tokenizers module integrates seamlessly with the geniml library for genomic ML:
```python
# Tokenize regions for geniml
from gtars.tokenizers import TreeTokenizer
import geniml
tokenizer = TreeTokenizer.from_bed_file("training_regions.bed")
tokens = [tokenizer.tokenize(r.chrom, r.start, r.end) for r in regions]
# Use tokens in geniml models
model = geniml.Model(vocab_size=tokenizer.vocab_size)
```
## Configuration Format
Tokenizer configuration files support YAML format:
```yaml
# tokenizer_config.yaml
type: tree
resolution: 1000 # Token resolution in base pairs
chromosomes:
- chr1
- chr2
- chr3
options:
overlap_handling: merge
gap_threshold: 100
```
## Performance Considerations
- TreeTokenizer uses efficient data structures for fast tokenization
- Batch tokenization is recommended for large datasets
- Pre-loading tokenizers reduces overhead for repeated operations
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the work | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixes | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation code | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |