Upload folder using huggingface_hub

README.md

@@ -1,3 +1,6 @@
+Updated README with the corrected file lists and without a summary section:
+
+````markdown
 ---
 pretty_name: genatator-segmentation-dataset
 configs:
@@ -5,7 +8,9 @@ configs:
   data_files:
   - split: train
     path:
-    - "train-human/data.parquet"
+    - "train-human/part-00001.parquet"
+    - "train-human/part-00002.parquet"
+    - "train-human/part-00003.parquet"
 
 - config_name: train-multi-specie
   data_files:
@@ -24,33 +29,50 @@ configs:
     - "train-multi-specie/part-00011.parquet"
     - "train-multi-specie/part-00012.parquet"
     - "train-multi-specie/part-00013.parquet"
+    - "train-multi-specie/part-00014.parquet"
+    - "train-multi-specie/part-00015.parquet"
+    - "train-multi-specie/part-00016.parquet"
+    - "train-multi-specie/part-00017.parquet"
+    - "train-multi-specie/part-00018.parquet"
+    - "train-multi-specie/part-00019.parquet"
+    - "train-multi-specie/part-00020.parquet"
+    - "train-multi-specie/part-00021.parquet"
+    - "train-multi-specie/part-00022.parquet"
+    - "train-multi-specie/part-00023.parquet"
+    - "train-multi-specie/part-00024.parquet"
+    - "train-multi-specie/part-00025.parquet"
+    - "train-multi-specie/part-00026.parquet"
+    - "train-multi-specie/part-00027.parquet"
+    - "train-multi-specie/part-00028.parquet"
+    - "train-multi-specie/part-00029.parquet"
 
 - config_name: val-human
   data_files:
   - split: validation
     path:
     - "val-human/data.parquet"
-
-- config_name: test-human-complete
-  data_files:
-  - split: test
-    path:
-    - "test-human-complete/data.parquet"
 ---
 
-# Ab inition gene segmentation benchmark (GENATATORs)
+# Ab initio gene segmentation benchmark (GENATATORs)
 
 ## Overview
 
-`genatator-segmentation-dataset` is a nucleotide-level gene segmentation dataset designed for training and evaluating DNA language models and related sequence models on transcript structure prediction. The dataset targets biologically detailed reconstruction of transcript architecture. In particular, it utilizes nucleotide-resolution labels describing the internal organization of transcripts, including 5' untranslated region (5' UTR), exon, intron, 3' untranslated region (3' UTR), and coding sequence (CDS).
+`genatator-segmentation-dataset` is a nucleotide-level gene segmentation dataset designed for training and evaluating DNA language models and related sequence models on transcript structure prediction. The dataset targets biologically detailed reconstruction of transcript architecture and supports benchmarking in the context of *ab initio* gene annotation. Each example represents one annotated transcript and provides nucleotide-resolution labels describing transcript organization, including 5' untranslated region (5' UTR), exon, intron, 3' untranslated region (3' UTR), and coding sequence (CDS).
 
-The dataset was created for benchmarking and developing models for transcript segmentation in the context of *ab initio* gene annotation. It supports both human-only and multispecies training setups, as well as held-out validation and test evaluation.
+The repository contains three configurations:
 
-Each sample contains exactly three fields:
+- `train-human`
+- `train-multi-specie`
+- `val-human`
+
+In the current version of the dataset, all configurations retain **all annotated transcripts for all genes**. No transcript-level truncation is applied. This design supports transcript-level segmentation studies while also enabling gene-level evaluation in multi-isoform settings; however, in many practical training and benchmarking scenarios, researchers may restrict the dataset to a single representative transcript per gene using the `status` field.
+
+Each sample contains exactly four fields:
 
 - `dna_sequence`
 - `labels`
 - `metadata`
+- `status`
 
 ## Intended use
 
@@ -61,46 +83,56 @@ This dataset is intended for:
 - benchmarking nucleotide-level and gene-structure-aware segmentation methods
 - evaluating generalization from human-only to multispecies training
 - studying segmentation of both protein-coding and long non-coding transcripts
+- evaluating models in settings where multiple transcript isoforms per gene are retained
 
-It is particularly suitable for methods that operate on long genomic or transcript-derived sequences and output per-nucleotide labels.
+It is particularly suitable for methods that operate on long genomic or transcript-derived sequences and produce per-nucleotide predictions.
 
 ## Dataset configurations
 
-The repository contains four configurations.
+The repository contains three configurations.
 
 | Config | Split | Description |
 |---|---|---|
-| `train-human` | `train` | Human-only training dataset. Sequence length is capped at 250 kb. One transcript per gene is retained, chosen as the isoform with the longest cumulative exon length. |
-| `train-multi-specie` | `train` | Multispecies training dataset spanning human and additional mammalian assemblies. Sequence length is capped at 250 kb. One transcript per gene is retained, chosen as the isoform with the longest cumulative exon length. |
-| `val-human` | `validation` | Human validation dataset. Sequence length is capped at 250 kb. One transcript per gene is retained, chosen as the isoform with the longest cumulative exon length. |
-| `test-human-complete` | `test` | Human test dataset containing full transcript sequences from chromosome 20 of the T2T human genome, with no truncation by length and with all annotated transcripts of each gene retained. |
+| `train-human` | `train` | Human training dataset containing all annotated transcripts for all genes. |
+| `train-multi-specie` | `train` | Multispecies training dataset containing all annotated transcripts for all genes across human and additional mammalian assemblies. |
+| `val-human` | `validation` | Human validation dataset containing all annotated transcripts for all genes. This configuration also provides chromosome 20 examples used for final model evaluation and gene-level metric calculation. |
 
-The dataset follows different transcript selection rules for training, validation, and test.
+## Dataset organization and evaluation protocol
 
-- `test-human-complete` contains all transcripts of each gene from chromosome 20 of the T2T human genome.
-- `train-human`, `train-multi-specie`, and `val-human` contain only one transcript per gene, selected as the transcript with the longest cumulative exon length.
+All three configurations retain the full set of annotated transcripts for each gene. Consequently, genes with multiple transcript isoforms are represented by multiple rows in the dataset.
 
-Human chromosomes are partitioned so that chromosomes 8, 20, and 21 are held out from human training. In this setup:
+Human chromosomes are partitioned so that chromosomes 8, 20, and 21 are excluded from human training. In this setup:
 
-- human training data exclude chromosomes 8, 20, and 21
-- human validation data are drawn from held-out human chromosomes
-- the full-length transcript version of human chromosome 20 is used for the `test-human-complete` dataset
+- `train-human` excludes human chromosomes 8, 20, and 21
+- `val-human` contains held-out human chromosomes, including chromosome 20
+- chromosome 20 from `val-human` is used for final model evaluation and for calculation of the gene-level metric
+- gene-level evaluation on chromosome 20 uses **all available transcripts per gene**, rather than a single representative isoform
 
 For the multispecies dataset:
 
-- all chromosomes are included for non-human species
-- human examples follow the same held-out chromosome policy described above
+- human examples follow the same held-out chromosome policy
+- non-human species are included according to the multispecies construction protocol used in dataset generation
+
+## Sequence length policy
+
+All transcripts are stored in full length.
+
+- No transcript is truncated.
+- No maximum transcript-length cap is applied in the released dataset.
+- Each `dna_sequence` and its aligned `labels` correspond to the complete sequence of the represented transcript.
+
+This is important for long-context modeling, complete transcript reconstruction, and biologically rigorous evaluation of exon-intron architecture.
 
 ## Data schema
 
-Each row has exactly three columns.
+Each row has exactly four columns.
 
 ### `dna_sequence`
 
-A string containing the DNA sequence for the example.
+A string containing the DNA sequence for the transcript.
 
 - Type: `string`
-- Alphabet: uppercase DNA characters (only A, T, C and G)
+- Alphabet: uppercase DNA characters (`A`, `T`, `C`, `G`)
 
 ### `labels`
 
@@ -108,7 +140,7 @@ A nested array of nucleotide-level target annotations aligned to `dna_sequence`.
 
 - Type: nested numeric array
 - Shape: sequence-length by class-dimension
-- Interpretation: per-nucleotide multilabel or multiclass segmentation targets used for transcript structure prediction
+- Interpretation: per-nucleotide segmentation targets used for transcript structure prediction
 
 The target class order is:
 
@@ -121,59 +153,89 @@ The target class order is:
 A compact string encoding transcript-level annotation in the following format:
 
 ```text
-<type>|<gene_name>|<transcript_name>|<strand>|<start>:<end>|<genome>
+<transcript_id>|<gene_id>|<transcript_type>|<strand>|<genome>|<chrom>|<start>:<end>
 ```
 
-## Metadata fields
+This schema is identical across all dataset configurations.
 
-The `metadata` field contains biologically interpretable attributes packed into a single string. The meaning of each component is described below.
+### `status`
 
-### 1. `type`
+A binary indicator identifying the representative transcript within a gene.
 
-Transcript class.
+* Type: integer
+* Typical values: `0` or `1`
 
-Typical values include:
+Interpretation:
 
-* `mRNA`
-* `lnc_RNA`
+* `status = 1` marks the transcript selected as the representative isoform for its gene
+* for protein-coding transcripts, this corresponds to the transcript with the **longest coding region**
+* for lncRNA transcripts, this corresponds to the transcript with the **longest cumulative exon length**
+* `status = 0` denotes all other transcripts of the same gene
 
-This field indicates whether the transcript is protein-coding or long non-coding.
+This field is useful for researchers who wish to run training with a single transcript per gene while retaining access to the complete multi-isoform dataset. In such cases, one should restrict the data to rows with `status == 1`.
 
-### 2. `gene_name`
+## Metadata fields
 
-Gene identifier or gene name associated with the transcript.
+The `metadata` column contains biologically interpretable attributes packed into a single string.
 
-Typical values look like:
+### 1. `transcript_id`
 
-* `gene-LOC124908100`
+Transcript identifier.
 
-This field identifies the parent gene for the transcript.
+Typical values may correspond to transcript accessions or annotation-specific transcript names.
 
-### 3. `transcript_name`
+This field identifies the specific transcript isoform represented by the row.
 
-Transcript identifier.
+### 2. `gene_id`
+
+Gene identifier associated with the transcript.
+
+Typical values may correspond to reference gene identifiers or annotation-derived gene names.
+
+This field identifies the parent gene of the transcript.
+
+### 3. `transcript_type`
+
+Transcript class.
 
-Typical values look like:
+Typical values include:
 
-* `rna-XR_007089385.1`
-* transcript accession-like names from reference annotations
+* `mRNA`
+* `lnc_RNA`
 
-This field identifies the specific transcript isoform represented by the example.
+This field indicates whether the transcript is protein-coding or long non-coding.
 
 ### 4. `strand`
 
 Genomic strand orientation.
 
-Allowed values are typically:
+Typical values are:
 
 * `+`
 * `-`
 
 This field indicates whether the transcript is encoded on the forward or reverse strand relative to the reference assembly.
 
-### 5. `start:end`
+### 5. `genome`
+
+Genome or assembly identifier associated with the example.
+
+Typical values include assembly accessions such as:
+
+* `GCF_009914755.1`
+* `GCF_000001635.26`
+
+This field is particularly important in the multispecies dataset.
+
+### 6. `chrom`
+
+Chromosome or reference sequence identifier on which the transcript is located.
 
-Genomic coordinate interval associated with the example.
+This field specifies the genomic contig or chromosome associated with the transcript.
+
+### 7. `start:end`
+
+Genomic coordinate interval associated with the transcript.
 
 Example:
 
@@ -181,21 +243,24 @@ Example:
 23090370:23092686
 ```
 
-This field stores the coordinate span as:
+This field stores the genomic span as:
 
-* `start`: integer genomic start position
-* `end`: integer genomic end position
+* `start`: 1-based genomic start coordinate
+* `end`: genomic end coordinate
 
-### 6. `genome`
+## Representative-transcript filtering with `status`
 
-Genome or assembly identifier associated with the example.
+Although the dataset retains all transcript isoforms, some training or evaluation protocols may require one transcript per gene. The `status` column was introduced precisely for this purpose.
 
-Typical values are assembly accessions such as:
+Recommended use:
 
-* `GCF_009914755.1`
-* `GCF_000001635.26`
+* use the full dataset when studying transcript diversity, isoform-aware evaluation, or gene-level metrics
+* filter to `status == 1` when a single representative transcript per gene is required
+
+Selection rule for `status == 1`:
 
-This field is especially relevant for the multispecies dataset, where it identifies the source assembly of the transcript example.
+* protein-coding genes: transcript with the longest coding region
+* lncRNA genes: transcript with the longest cumulative exon length
 
 ## Multispecies training dataset
 
@@ -253,7 +318,6 @@ from datasets import load_dataset
 train_human = load_dataset("shmelev/genatator-segmentation-dataset", "train-human")["train"]
 train_multi = load_dataset("shmelev/genatator-segmentation-dataset", "train-multi-specie")["train"]
 val_human = load_dataset("shmelev/genatator-segmentation-dataset", "val-human")["validation"]
-test_human = load_dataset("shmelev/genatator-segmentation-dataset", "test-human-complete")["test"]
 ```
 
 Access one example:
@@ -263,8 +327,20 @@ sample = train_human[0]
 print(sample["dna_sequence"])
 print(sample["labels"])
 print(sample["metadata"])
+print(sample["status"])
 ```
 
-## Summary
+Filter to one representative transcript per gene:
+
+```python
+representative_only = train_human.filter(lambda x: x["status"] == 1)
+```
+
+## Recommended evaluation usage
+
+For final model evaluation and for calculation of the gene-level metric:
+
+* use chromosome 20 from `val-human`
+* retain all available transcripts per gene
+* compare predictions against the complete isoform set for each gene
 
-`genatator-segmentation-dataset` is a long-context, nucleotide-level transcript segmentation dataset for DNA language models and related genomic sequence models. It includes human-only and multispecies training resources, a human validation set, and a full-length human test set, all formatted for direct use with modern machine learning pipelines.

train-human/part-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1eb444ea291fb79cc63c46c1b7cccb953cdc037a3d403f62b50d28245449084c
+size 400442227

train-human/part-00002.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54f933ef4708a3543e9f14c3e936fd5f961456d2e366c1a023ba033776ee2b41
+size 605030095

train-human/part-00003.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:53260b594c81a088dac3614ac59a6cc85dfa167a28da6639ead8f0a0f462a4c1
+size 572547632

train-multi-specie/part-00001.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:776c3b568d29c4485221638aada86bc3e74707c6b04e4a08d3236d5541cebadb
-size 255216895
+oid sha256:71f947c54e7da89174fe4e7b9c22180e39f5dfe3e6852f92a3de3a455a526215
+size 597663398

train-multi-specie/part-00002.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:65f1ba8c40df8cc23335cbd501e115c712d6a6c8460add46d05b969e74a1a23a
-size 257528241
+oid sha256:5fa6d262973551e6ebec6eac39f85823bfd1e63fc452a597f683f3cab3b9657c
+size 625895518

train-multi-specie/part-00003.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5ddb80a08e7e931ecc3cff23df134e3fd1721657d9f162f26f738b5cb6c61ed2
-size 292795898
+oid sha256:68e49e72bc4f02e2b23c74ba940d288eac534ea5f46e3431cd56ed98103d937b
+size 447158408

train-multi-specie/part-00004.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ab087f9b7e40dfc558065a8816055b05a7b4a70882205747b5b05e3d72c3fe0a
-size 219343258
+oid sha256:a3377d0d60e6af1c47d1740ea602241d6d9d9737053114ebc91deb015ad7b88b
+size 496018576

train-multi-specie/part-00005.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:24d45bacb591d13359fde76ecf0ee98d221e1619c83cf86c6d7bad6ac80f400c
-size 285652767
+oid sha256:4548246a5b36c79c7c01d7a7b5171e56383c0772a57ee2403616407f7d2d98ae
+size 438020088

train-multi-specie/part-00006.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c402f349e71ce742c7eecfa2d228b497d8b85ad6cba6847c51aa830693768031
-size 349053314
+oid sha256:c0b4b2d68da5410ba54d05e797d73767e94ea8a2f149ff9b13dd80848e4c8bea
+size 638035247

train-multi-specie/part-00007.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7c783858c0879074df98d6c82fc9910c7cd498515335914c188d50d158d350ef
-size 242536152
+oid sha256:ca08ca412c1e74486be047d1131627cbca3893270e33cf2a4b79c01e4bf4654a
+size 625778082

train-multi-specie/part-00008.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:07a3d7492acb65d49b7ad0cdb7fa0debaa0b35a88ed87d6c61a9e4b6ec7fd6b5
-size 167473428
+oid sha256:0775c257f2b5318e39b258c6b8c7a30d216bafd51be8441fbdf6915cabf9cc7c
+size 449042226

train-multi-specie/part-00009.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a3bd0430908ed638a7853a2e7c2d43c74ff99b855f28d8ce264d075843c64dd8
-size 484941957
+oid sha256:32f0970bc5e9f0ce58cf3874b08317bdc7a53c90e59ba273b16e670022bc23ab
+size 488334782

train-multi-specie/part-00010.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:97d0dceff9da3f778293aaeea8207497d71d6fd6c4385eeca321066a54024013
-size 452647842
+oid sha256:0f2112bcaf53db0b18fe6129e97affd36a50d1ffa899553153a3a4ebf8b477a4
+size 299985745

train-multi-specie/part-00011.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4aa3ead8d3afac2be77dd0d549873eba1663e1ab151f328de1097ee22158f3d6
-size 554848780
+oid sha256:cf2362591b24b1c900e47001b7fa9451f8a1c3f027f1c7eae11eac52a8615246
+size 283620455

train-multi-specie/part-00012.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0faadfeb439d24e9d10702a1a708cd4a054572659005fcce957d9d67a881503a
-size 482879736
+oid sha256:cf721ab5f1771047ac93b61618b7c277430eb50cfb672274f7db541e825bbb39
+size 308425385

train-multi-specie/part-00013.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fe906a8251bc238297e4a2960c4a2d00dfc1b6fef5a1e14ccd006c54d154c041
-size 217706485
+oid sha256:0f18769dd0d10f3a08f5deef5ca89e63ba1e9933a6c06b01b53a37ee3243344b
+size 313520857

train-multi-specie/part-00014.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae8a983e89e93d2cb8212d0eb64df0303cba30627ff240789a856851de6f4547
+size 223770801

train-multi-specie/part-00015.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:36d21f438e206c46e6d20160fd62dd35331e238a019e73bc752f847cae918c58
+size 234867916

train-multi-specie/part-00016.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5788fb45cd9b8f17255efb38845e04e4fd6682c2c7cd41f1f9c883b304dd2bf
+size 243604235

train-multi-specie/part-00017.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0365caaf8f568a2d70f5fdb5ff7f8c877058f6f68105fcd97b7689e396e2ded2
+size 299158476

train-multi-specie/part-00018.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:65166383357ada494d8f59170971a61c9636b63e272b99a51a7ce61db04bd51a
+size 214411400

train-multi-specie/part-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd7bec21b385fccbb0dea4688d6c7c8f852314a39ca1460b36898b754c264eb7
+size 180597346

train-multi-specie/part-00020.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:886aef78f627f3dcd5cd5f87ce5bb10566ed7d4c918408536a9e083381b255bd
+size 223418891

train-multi-specie/part-00021.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e2a2d588dfe0c729cb597332050ceb556f6d361613ec05df905b2e4576d3bf11
+size 280526952

train-multi-specie/part-00022.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:42bcbf10e18cc0a1c0945fa0616efbd98258efe09b4319ea63bdbefa1dd764ac
+size 293892826

train-multi-specie/part-00023.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a2349cd9dd59eaa5cf6e3f7b1f87618e75fc2a096a6c62efb3e4413130e29dc5
+size 457879258

train-multi-specie/part-00024.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2d3c749e21a998289eaf376d21d8d1fc029de52973ab9097ebc26d9975ee5869
+size 540326778

train-multi-specie/part-00025.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:18b55d7b9d66ab8921b44e35f58ee6bd508d9650d784902a2903506d3cf5cdb7
+size 342620595

train-multi-specie/part-00026.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:46de861ca345de3f59db26a64acbd477612cc90973a1e5301aba22bdd58b821c
+size 200615672

train-multi-specie/part-00027.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ec1e8ba6ea1f14f21b47aefb0071d6ce207aee5ff03fb6d18c85e3d9b463b5b
+size 212130648

train-multi-specie/part-00028.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd3f3045bd1f974c5f72f72fb7752474499c91d853b5529cb5b3ed8af05cdfd8
+size 205619573

train-multi-specie/part-00029.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4eb7747a50e8a73b5b83e377a1502943abf236e8e8d5c0d1f8008dba1bb0bca3
+size 541510536

val-human/data.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2364adfa677c9fb2d13ad0962567170f4844fcee1291f3553e5b89d869f9c56e
-size 25075170
+oid sha256:9ebcb0d4aee976f30b1adef09f1ea0f70ef29434d1c63497ef276a8367b11fe1
+size 129492405