mgyigit commited on
Commit
4f747e9
·
verified ·
1 Parent(s): 126b728

Update src/about.py

Browse files
Files changed (1) hide show
  1. src/about.py +3 -5
src/about.py CHANGED
@@ -91,8 +91,8 @@ LLM_BENCHMARKS_TEXT = f"""
91
  EVALUATION_QUEUE_TEXT = """
92
  # **Benchmarking your own representation model**
93
  ## To run the benchmarks, the following representation vectors need to be generated:
94
- - For benchmarks 1, 2, and 3 (similarity, function, and family), you will need to generate representation vectors for all human proteins. The amino acid sequences for canonical isoforms of human proteins can be found [here](https://drive.google.com/file/d/1wXF2lmj4ZTahMrl66QpYM2TvHmbcIL6b/view?usp=sharing).
95
- - For benchmark 4 (affinity), representation vectors will need to be generated for the samples in the SKEMPI dataset, which can be accessed [here](https://drive.google.com/file/d/1m5jssC0RMsiFT_w-Ykh629Pw_An3PInI/view?usp=sharing).
96
 
97
  ## Format of the both protein representation files:
98
  1. Each row corresponds to the representation vector of a particular protein.
@@ -133,11 +133,9 @@ family_prediction_dataset_options = ["nc", "uc50", "uc30", "mm15"]
133
 
134
  benchmark_specific_metrics = {
135
  'similarity': ['sparse_MF_correlation', 'sparse_BP_correlation', 'sparse_CC_correlation', 'sparse_Ave_correlation',
136
- 'sparse_MF_pvalue', 'sparse_BP_pvalue', 'sparse_CC_pvalue', 'sparse_Ave_pvalue',
137
  '200_MF_correlation', '200_BP_correlation', '200_CC_correlation', '200_Ave_correlation',
138
- '200_MF_pvalue', '200_BP_pvalue', '200_CC_pvalue', '200_Ave_pvalue',
139
  '500_MF_correlation', '500_BP_correlation', '500_CC_correlation', '500_Ave_correlation',
140
- '500_MF_pvalue', '500_BP_pvalue', '500_CC_pvalue', '500_Ave_pvalue',],
141
  'function': {
142
  'aspect_types': ['MF', 'BP', 'CC'],
143
  'dataset_types': ['accuracy', 'F1', 'precision', 'recall']
 
91
  EVALUATION_QUEUE_TEXT = """
92
  # **Benchmarking your own representation model**
93
  ## To run the benchmarks, the following representation vectors need to be generated:
94
+ - For benchmarks 1, 2, and 3 (similarity, function, and family), you will need to generate representation vectors for all human proteins. The amino acid sequences for canonical isoforms of human proteins can be found [here](https://drive.google.com/file/d/1wXF2lmj4ZTahMrl66QpYM2TvHmbcIL6b/view?usp=sharing).
95
+ - For benchmark 4 (affinity), representation vectors will need to be generated for the samples in the SKEMPI dataset, which can be accessed [here](https://drive.google.com/file/d/1m5jssC0RMsiFT_w-Ykh629Pw_An3PInI/view?usp=sharing).
96
 
97
  ## Format of the both protein representation files:
98
  1. Each row corresponds to the representation vector of a particular protein.
 
133
 
134
  benchmark_specific_metrics = {
135
  'similarity': ['sparse_MF_correlation', 'sparse_BP_correlation', 'sparse_CC_correlation', 'sparse_Ave_correlation',
 
136
  '200_MF_correlation', '200_BP_correlation', '200_CC_correlation', '200_Ave_correlation',
 
137
  '500_MF_correlation', '500_BP_correlation', '500_CC_correlation', '500_Ave_correlation',
138
+ ],
139
  'function': {
140
  'aspect_types': ['MF', 'BP', 'CC'],
141
  'dataset_types': ['accuracy', 'F1', 'precision', 'recall']