Hugging Face's logo

Spaces:
HUBioDataLab
/
PROBE
Running

App Files Community
PROBE
File size: 5,316 Bytes 
dd49f8a
 
 
4db2d24
 
 
 
dd49f8a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a2e6203
 
dd49f8a
a2e6203
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
 
import yaml
import pandas as pd
import tqdm
from . import semantic_similarity_infer as ssi
from . import target_family_classifier as tfc
from . import function_predictor as fp
from . import binding_affinity_estimator as bae

print("\n\nPROBE (Protein RepresentatiOn Benchmark) run is started...\n\n")

with open('probe_config.yaml') as f:
	args = yaml.load(f, Loader=yaml.FullLoader)

if args["benchmark"] not in ["similarity","family","function","affinity","all"]:
        parser.error('At least one benchmark type should be selected')

print(args)

def load_representation(multi_col_representation_vector_file_path):
    multi_col_representation_vector = pd.read_csv(multi_col_representation_vector_file_path)
    vals = multi_col_representation_vector.iloc[:,1:(len(multi_col_representation_vector.columns))]
    original_values_as_df = pd.DataFrame({'Entry': pd.Series([], dtype='str'),'Vector': pd.Series([], dtype='object')})
    for index, row in tqdm.tqdm(vals.iterrows(), total = len(vals)):
        list_of_floats = [float(item) for item in list(row)]
        original_values_as_df.loc[index] = [multi_col_representation_vector.iloc[index]['Entry']] + [list_of_floats]
    return original_values_as_df

if args["benchmark"] in  ["similarity","function","all"]:
    print("\nRepresentation vectors are loading...\n")
    representation_dataframe = load_representation(args["representation_file_human"])
 
if args["benchmark"] in  ["similarity","all"]:
    print("\nSemantic similarity Inference Benchmark is running...\n")
    ssi.representation_dataframe = representation_dataframe
    ssi.representation_name = args["representation_name"]
    ssi.protein_names = ssi.representation_dataframe['Entry'].tolist()
    ssi.similarity_tasks = args["similarity_tasks"]
    ssi.detailed_output = args["detailed_output"]
    ssi.calculate_all_correlations()
if args["benchmark"] in  ["function","all"]:
    print("\n\nOntology-based protein function prediction benchmark is running...\n")
    fp.aspect_type = args["function_prediction_aspect"]
    fp.dataset_type = args["function_prediction_dataset"]
    fp.representation_dataframe = representation_dataframe
    fp.representation_name = args["representation_name"]
    fp.detailed_output = args["detailed_output"]
    fp.pred_output()
if args["benchmark"] in  ["family","all"]:
    print("\n\nDrug target protein family classification benchmark is running...\n")
    tfc.representation_path = args["representation_file_human"]
    tfc.representation_name = args["representation_name"]
    tfc.detailed_output = args["detailed_output"]
    for dataset in args["family_prediction_dataset"]:
	    tfc.score_protein_rep(dataset)
if args["benchmark"] in  ["affinity","all"]:
    print("\n\nProtein-protein binding affinity estimation benchmark is running...\n")
    bae.skempi_vectors_path = args["representation_file_affinity"]
    bae.representation_name = args["representation_name"]
    bae.predict_affinities_and_report_results()
print("\n\nPROBE (Protein RepresentatiOn Benchmark) run is finished...\n")

def run_probe(benchmarks, representation_name, representation_file_human, representation_file_affinity, similarity_tasks=["Sparse","200","500"], function_prediction_aspec="All_Aspects", function_prediction_dataset="All_Data_Sets", family_prediction_dataset=["nc","uc50","uc30","mm15"], detailed_output=False):
    print("\n\nPROBE (Protein RepresentatiOn Benchmark) run is started...\n\n")

    if any(item in ['similarity', 'function', 'family', 'all'] for item in benchmarks):
        print("\nRepresentation vectors are loading...\n")
        human_representation_dataframe = load_representation(representation_file_human)

    if "similarity" in benchmarks:
        print("\nSemantic similarity Inference Benchmark is running...\n")
        ssi.representation_dataframe = human_representation_dataframe
        ssi.representation_name = representation_name
        ssi.protein_names = ssi.representation_dataframe['Entry'].tolist()
        ssi.similarity_tasks = similarity_tasks
        ssi.detailed_output = detailed_output
        ssi.calculate_all_correlations()

    if "function" in benchmarks:
        print("\n\nOntology-based protein function prediction benchmark is running...\n")
        fp.aspect_type = function_prediction_aspect
        fp.dataset_type = function_prediction_dataset
        fp.representation_dataframe = human_representation_dataframe
        fp.representation_name = representation_name
        fp.detailed_output = detailed_output
        fp.pred_output()

    if "family" in benchmarks:
        print("\n\nDrug target protein family classification benchmark is running...\n")
        tfc.representation_path = representation_file_human
        tfc.representation_name = representation_name
        tfc.detailed_output = detailed_output
        for dataset in family_prediction_dataset:
            tfc.score_protein_rep(dataset)

    if "affinity" in benchmarks:
        print("\n\nProtein-protein binding affinity estimation benchmark is running...\n")
        bae.skempi_vectors_path = representation_file_affinity
        bae.representation_name = representation_name
        bae.predict_affinities_and_report_results()

    print("\n\nPROBE (Protein RepresentatiOn Benchmark) run is finished...\n")
    return 0