Merge origin/master: integrate new docs and refget updates

Author: nsheff

docs/bbconf/changelog.md

@@ -2,6 +2,12 @@
 
 This project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html) and [Keep a Changelog](https://keepachangelog.com/en/1.0.0/) format.
 
+
+### [0.13.0] - 2025-12-01
+### Added:
+- UMAP calculation
+
+
 ### [0.12.0] - 2025-09-11
 ### Added:
 - New qdrant semantic search

docs/bedbase/user/bbclient.md

@@ -53,7 +53,7 @@ Full usage documentation is available in the Usage documentation is available [B
 RegionSet is a Python/Rust/R representation of a BED file. It allows user to compute identifiers, save bed files, 
 iterate through regions, and perform other operations on the BED file.
 
-How to install and use RegionSet in Python is described in the [RegionSet documentation](../../gtars/python/regionset.md). 
+How to install and use RegionSet in Python is described in the [RegionSet documentation](../../gtars/models.md). 
 
 Quick example:
 

docs/bedbase/user/bed-analyzer.md

@@ -0,0 +1 @@
+🚧 This documentation is under active development and will be updated soon. 🚧

docs/bedbase/user/umap-visualizer.md

@@ -0,0 +1 @@
+🚧 This documentation is under active development and will be updated soon. 🚧

docs/bedhost/changelog.md

@@ -2,6 +2,14 @@
 
 This project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html) and [Keep a Changelog](https://keepachangelog.com/en/1.0.0/) format. 
 
+## [0.12.0] -- 2025-12-01
+### Added:
+- umap calculation [API]
+- BED-analyzer (qc), which is backed by gtars-wasm package [UI]
+- New improved home page [UI]
+- Added visualization of the UMAP, and interactive [UI]
+
+
 ## [0.11.0] -- 2025-09-11
 ### Added:
 - Added umap visualizations of bed embeddings

docs/citations.md

@@ -8,18 +8,26 @@ Components of BEDbase and geniml have been published independently. Here, we sho
 
 Thanks for citing us! If you use BEDbase, geniml, or their components in your research, here are papers you can cite.
 
-| If you use... | Please cite ... |
-|---------------|-----------------|
-| `region2vec` embeddings | Gharavi et al. (2021) *Bioinformatics* |
-| `bedspace` search and embeddings | Gharavi et al. (2024) *Bioengineering* |
+| If you use...                             | Please cite ... |
+|-------------------------------------------|-----------------|
+| `region2vec` embeddings                   | Gharavi et al. (2021) *Bioinformatics* |
+| `bedspace` search and embeddings          | Gharavi et al. (2024) *Bioengineering* |
 | `scEmbed` single-cell embedding framework | LeRoy et al. (2024) *NAR Genomics and Bioinformatics* |
-| `geniml` region set evaluations | Zheng et al. (2024) *NAR Genomics and Bioinformatics* |
-| `geniml hmm` module | Rymuza et al. (2024) *Nucleic Acis Research* |
-| `bedbase` database | Unpublished |
+| `geniml` region set evaluations           | Zheng et al. (2024) *NAR Genomics and Bioinformatics* |
+| `geniml hmm` module                       | Rymuza et al. (2024) *Nucleic Acis Research* |
+| `bedbase` database                        | Unpublished |
+| `tokenizers` module                       | LeRoy et al. (2025) *arXiv* |
+| `atacformer` foundation model             | LeRoy et al. (2025) *bioRxiv* |
 
 
 ## Full citation information for manuscripts
 
+- **LeRoy et al. (2025).** *Atacformer: A transformer-based foundation model for analysis and interpretation of ATAC-seq data*.
+  *bioRxiv*. [10.1101/2025.11.03.685753](https://doi.org/10.1101/2025.11.03.685753)
+
+- **LeRoy et al. (2025).** *Fast, memory-efficient genomic interval tokenizers for modern machine learning*.
+  *arXiv*. [10.48550/arXiv.2511.01555](https://doi.org/10.48550/arXiv.2511.01555) 
+
 - **Gharavi et al. (2024).** *Joint representation learning for retrieval and annotation of genomic interval sets*.
   *Bioengineering*. [10.3390/bioengineering11030263](http://dx.doi.org/10.3390/bioengineering11030263)
 

docs/geniml/tutorials/fine-tune-region2vec-model.md

@@ -5,7 +5,7 @@ Fine-tuning a model is a way to adapt a pre-trained model to a new task. For exa
 ## Get a pretrained model
 To begin, we need to get a pretrained model. We can get one from huggingface:
 ```python
-from geniml.region2vec import Region2VecExModel
+from geniml.region2vec.main import Region2VecExModel
 
 model = Region2VecExModel("databio/r2v-ChIP-atlas-hg38-v2")
 ```

docs/geniml/tutorials/train-region2vec.md

@@ -21,7 +21,7 @@ from multiprocessing import cpu_count
 
 from geniml.io import RegionSet
 from geniml.tokenization import TreeTokenizer
-from geniml.region2vec import Region2VecExModel
+from geniml.region2vec.main import Region2VecExModel
 from rich.progress import track
 
 

docs/geniml/tutorials/use-pretrained-region2vec-model.md

@@ -6,7 +6,7 @@ To use one of our pre-trained models, simply import the `Region2VecExModel` and
 
 ```python
 from geniml.io import Region
-from geniml.region2vec import Region2VecExModel
+from geniml.region2vec.main import Region2VecExModel
 
 model = Region2VecExModel("databio/r2v-encode-hg38")
 ```

docs/gtars/models.md

@@ -0,0 +1,134 @@
+# Models and Region Set objects in Gtars
+
+Gtars has multiple objects (structs/models) for representation of genomic regions and other related data. 
+
+### 🟢 Region
+
+Region is Python representation of a genomic region. e.g. `chr1:100-200` + additional information.
+
+#### Example
+
+=== "Python"
+    ```python
+    from gtars.models import Region
+
+    # Create a Region
+    genomic_region = Region(chr="chr1", 
+                             start=100, 
+                             end=200, 
+                             rest="peak1")
+    print(genomic_region)
+
+    ```
+
+=== "Rust"
+    ```rust
+    use gtars::models::Region;
+    
+    // Create a Region
+    let genomic_region: Region = Region { chr: "chr1".to_string(), 
+                                          start: 100, 
+                                          end: 200, 
+                                          rest: Some("peak1".to_string()) 
+                                        },
+    let identifier = genomic_region.digest();
+
+    println!("{:?}", identifier);
+    
+    ```
+
+
+### 🟢 RegionSet
+
+RegionSet is Python representation of a genomic region set, commonly named as BED file.
+
+
+#### Quick example
+Open BED file from URL and get its identifier.
+
+=== "Python"
+    ```python
+    
+    from gtars.models import RegionSet
+    
+    # Create a RegionSet from a url, or lcoal BED file.
+    rs = RegionSet("https://data2.bedbase.org/files/d/a/dafd661aa70590999e0ff9e1980217db.bed.gz")
+    
+    # Get identifier for the RegionSet
+    rs.identifier
+    
+    print(rs)
+
+    ```
+=== "Rust"
+    ```rust
+    use gtars::models::RegionSet;
+    
+    // Create a RegionSet from a url, or lcoal BED file.
+    let rs = RegionSet::try_from("https://data2.bedbase.org/files/d/a/dafd661aa70590999e0ff9e1980217db.bed.gz").unwrap();
+    
+    // Get identifier for the RegionSet
+    let id = rs.identifier();
+
+    println!("{:?}", rs);
+    ```
+
+=== "TypeScript"
+    
+    ❗ Note: This is test example and may require additional setup to run.
+    
+    ```typescript
+    import init from '@databio/gtars';
+    import { RegionSet } from '@databio/gtars';
+
+    init();
+
+    export type BedEntry1 = [string, number, number, string];
+    
+    // Define entries (regions)
+    export const entries1: BedEntry1[] = [
+      ['chr1', 100, 200, 'peak1'],
+      ['chr2', 150, 250, 'peak2'],
+      ['chr3', 300, 400, 'peak3'],
+    ];
+
+    // Create a Region
+    const rs = new RegionSet(entries1);
+
+    console.log(rs);
+
+    ```
+
+❗ Note: RegionSet can be created from a local file path, URL, or by passing a list (vector) or Region objects.
+
+#### Main commands in Python
+
+- Load a BED file from local path or URL
+```python
+rs = RegionSet("path/to/bedfile.bed")
+```
+- Get number of regions
+```python
+len(rs)
+```
+- Calculate mean reagion width
+```python
+rs.mean_region_width()
+```
+- Get last base pair location for each chromosome
+```python
+rs.get_max_end_per_ch()
+```
+- Get number of base pairs in the region set
+```python
+rs.get_nucleotide_length()
+```
+- Save the regionSet as a BED file
+```python
+rs.to_bed("path/to/save/bedfile.bed")
+rs.to_bed_gz("path/to/save/bedfile.bed.gz")  # gzipped
+```
+- Save the regionSet as a bigBed file
+```python
+rs.to_bigbed("path/to/save/bedfile.bb", chrom_sizes="path/to/chrom.sizes")
+```