etrop commited on
Commit
5f5c6f4
1 Parent(s): 80dbee5

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +16 -3
README.md CHANGED
@@ -9,26 +9,27 @@ license: cc-by-nc-sa-4.0
9
 
10
 
11
  ## Dataset Overview
12
- This dataset features the 7 evaluation tasks presented in the AgroNT (A Foundational Large Language Model for Edible Plant
13
  Genomes) paper. The tasks cover single output regression, multi output regression, binary classification, and multi-label classification which
14
  aim to provide a comprehensive plant genomics benchmark. Additionally, we provide results from in silico saturation mutagenesis analysis of sequences
15
  from the cassava genome, assessing the impact of >10 million mutations on gene expression levels and enhancer elements. See the ISM section
16
  below for details regarding the data from this analysis.
17
 
18
 
19
- | Task Name | # of Datasets(Species) | Task Type | Sequence Length (base pair) |
20
  | -------- | ------- | -------- | ------- |
21
  | Polyadenylation | 6 | Binary Classification | 400 |
22
  | Splice Site | 2 | Binary Classification | 398 |
23
  | LncRNA | 6 | Binary Classification | 101-6000 |
24
  | Promoter Strength | 2 | Single Variable Regression | 170 |
 
25
  | Chromatin Accessibility | 7 | Multi-label Classification | 1000 |
26
  | Gene Expression | 6 | Multi-Variable Regression | 6000 |
27
  | Enhancer Region | 1 | Binary Classification | 1000 |
28
 
29
 
30
  ## Dataset Sizes
31
- | Config Name | # Train Samples | # Validation Samples | # Test Samples |
32
  | -------- | ------- | -------- | ------- |
33
  |poly_a.arabidopsis_thaliana|170835|---|30384|
34
  |poly_a.oryza_sativa_indica_group|98139|---|16776|
@@ -46,6 +47,8 @@ below for details regarding the data from this analysis.
46
  |lncrna.s_bicolor|8654|---|734|
47
  |promoter_strength.leaf|58179|6825|7154|
48
  |promoter_strength.protoplast|61051|7162|7595|
 
 
49
  |gene_exp.glycine_max|47136|4803|4803|
50
  |gene_exp.oryza_sativa|31244|3702|3702|
51
  |gene_exp.solanum_lycopersicum|27321|3827|3827|
@@ -63,6 +66,16 @@ below for details regarding the data from this analysis.
63
  *** It is important to note that fine-tuning for lncrna was carried out using all datasets in a single training. The reason for this is that the datasets are small and combining
64
  them helped to improve learning.
65
 
 
 
 
 
 
 
 
 
 
 
66
  ## In Silico Saturation Mutagensis
67
  ### File structure for: ISM_Tables/Mesculenta_305_v6_PROseq_ISM_LOG2FC.txt.gz
68
  Intergenic enhancer regions based on Lozano et al. 2021 (https://pubmed.ncbi.nlm.nih.gov/34499719/) <br>
 
9
 
10
 
11
  ## Dataset Overview
12
+ This dataset features the 8 evaluation tasks presented in the AgroNT (A Foundational Large Language Model for Edible Plant
13
  Genomes) paper. The tasks cover single output regression, multi output regression, binary classification, and multi-label classification which
14
  aim to provide a comprehensive plant genomics benchmark. Additionally, we provide results from in silico saturation mutagenesis analysis of sequences
15
  from the cassava genome, assessing the impact of >10 million mutations on gene expression levels and enhancer elements. See the ISM section
16
  below for details regarding the data from this analysis.
17
 
18
 
19
+ | Name | # of Datasets(Species) | Task Type | Sequence Length (base pair) |
20
  | -------- | ------- | -------- | ------- |
21
  | Polyadenylation | 6 | Binary Classification | 400 |
22
  | Splice Site | 2 | Binary Classification | 398 |
23
  | LncRNA | 6 | Binary Classification | 101-6000 |
24
  | Promoter Strength | 2 | Single Variable Regression | 170 |
25
+ | Terminator Strength | 2 | Single Variable Regression | 170 |
26
  | Chromatin Accessibility | 7 | Multi-label Classification | 1000 |
27
  | Gene Expression | 6 | Multi-Variable Regression | 6000 |
28
  | Enhancer Region | 1 | Binary Classification | 1000 |
29
 
30
 
31
  ## Dataset Sizes
32
+ | Task Name | # Train Samples | # Validation Samples | # Test Samples |
33
  | -------- | ------- | -------- | ------- |
34
  |poly_a.arabidopsis_thaliana|170835|---|30384|
35
  |poly_a.oryza_sativa_indica_group|98139|---|16776|
 
47
  |lncrna.s_bicolor|8654|---|734|
48
  |promoter_strength.leaf|58179|6825|7154|
49
  |promoter_strength.protoplast|61051|7162|7595|
50
+ |terminator_strength.leaf|43294|5309|4806|
51
+ |terminator_strength.protoplast|43289|5309|4811|
52
  |gene_exp.glycine_max|47136|4803|4803|
53
  |gene_exp.oryza_sativa|31244|3702|3702|
54
  |gene_exp.solanum_lycopersicum|27321|3827|3827|
 
66
  *** It is important to note that fine-tuning for lncrna was carried out using all datasets in a single training. The reason for this is that the datasets are small and combining
67
  them helped to improve learning.
68
 
69
+ ## Example Usage
70
+ ```python
71
+ from datasets import load_dataset
72
+
73
+ task_name='terminator_strength.protoplast' # one of the task names from the above table
74
+
75
+ dataset = load_dataset("InstaDeepAI/plant-genomic-benchmark",task_name=task_name)
76
+
77
+ ```
78
+
79
  ## In Silico Saturation Mutagensis
80
  ### File structure for: ISM_Tables/Mesculenta_305_v6_PROseq_ISM_LOG2FC.txt.gz
81
  Intergenic enhancer regions based on Lozano et al. 2021 (https://pubmed.ncbi.nlm.nih.gov/34499719/) <br>