Upload app.py

app.py

@@ -863,36 +863,36 @@ To predict interactions/binding affinities of a single target against a library
                                 "Alternatively, enter a Uniprot ID or gene symbol with organism and click Query for "
                                 "the sequence."
                             )
-                            with gr.Row():
-                                target_input_type = gr.Dropdown(
-                                    label='Step 1. Select Target Input Type and Input',
-                                    choices=['Sequence', 'UniProt ID', 'Gene symbol'],
-                                    info='Enter (paste) a FASTA string below manually or upload a FASTA file.',
-                                    value='Sequence',
-                                    scale=4, interactive=True
-                                )
-                                target_id = gr.Textbox(show_label=False, visible=False,
-                                                       interactive=True, scale=4,
-                                                       info='Query a sequence on UniProt with a UniProt ID.')
-                                target_gene = gr.Textbox(
-                                    show_label=False, visible=False,
-                                    interactive=True, scale=4,
-                                    info='Query a sequence on UniProt with a gene symbol.')
-                                target_organism = gr.Textbox(
-                                    info='Organism scientific name (default: Homo sapiens).',
-                                    placeholder='Homo sapiens', show_label=False,
-                                    visible=False, interactive=True, scale=4, )
+                            target_input_type = gr.Dropdown(
+                                label='Step 1. Select Target Input Type and Input',
+                                choices=['Sequence', 'UniProt ID', 'Gene symbol'],
+                                info='Enter (paste) a FASTA string below manually or upload a FASTA file.',
+                                value='Sequence',
+                                scale=4, interactive=True
+                            )
 
                     with gr.Row():
+                        target_id = gr.Textbox(show_label=False, visible=False,
+                                               interactive=True, scale=4,
+                                               info='Enter a UniProt ID and query.')
+                        target_gene = gr.Textbox(
+                            show_label=False, visible=False,
+                            interactive=True, scale=4,
+                            info='Enter a gene symbol and query.')
+                        target_organism = gr.Textbox(
+                            info='Organism scientific name (default: Homo sapiens).',
+                            placeholder='Homo sapiens', show_label=False,
+                            visible=False, interactive=True, scale=4, )
                         target_upload_btn = gr.UploadButton(label='Upload a FASTA file', type='binary',
                                                             visible=True, variant='primary',
                                                             size='lg')
+                        target_paste_markdown = gr.Button(value='Or paste your sequence below', visible=True)
                         target_query_btn = gr.Button(value='Query the sequence', variant='primary',
-                                                     visible=False)
+                                                     visible=False, scale=4)
                     # with gr.Row():
                     #     example_uniprot = gr.Button(value='Example: Q16539', elem_classes='example', visible=False)
                     #     example_gene = gr.Button(value='Example: MAPK14', elem_classes='example', visible=False)
-                    example_fasta = gr.Button(value='Example: Human MAPK14', elem_classes='example')
+                    example_fasta = gr.Button(value='Example: MAPK14 (Q16539)', elem_classes='example')
                     target_fasta = gr.Code(label='Input or Display FASTA', interactive=True, lines=5)
                     # with gr.Row():
                     #     with gr.Column():
@@ -980,7 +980,7 @@ To predict interactions/binding affinities of a single target against a library
 
         with gr.TabItem(label='Target protein identification', id=1):
             gr.Markdown('''
-# <center>DeepSEQreen Target Protein Identification</center>
+# <center>Target Protein Identification</center>
 
 <center>
 To predict interactions/binding affinities of a single compound against a library of protein targets.
@@ -1015,7 +1015,7 @@ To predict interactions/binding affinities of a single compound against a librar
                                 "If the proteins in the target library of interest all belong to the same protein "
                                 "family, manually selecting the family is supported."
                             )
-                            target_identify_target_family = gr.Dropdown(choices=list(TARGET_FAMILY_MAP.keys()),
+                            target_identify_target_family = gr.Dropdown(choices=['General'],
                                                                         value='General',
                                                                         label='Step 2. Select Target Protein Family ('
                                                                               'Optional)')
@@ -1042,7 +1042,7 @@ To predict interactions/binding affinities of a single compound against a librar
                         with gr.Column():
                             HelpTip(
                                 "Interaction prediction provides you binding probability score between the target of "
-                                "interest and each compound in the library,"
+                                "interest and each compound in the library, "
                                 "while affinity prediction directly estimates their binding strength measured using "
                                 "IC50."
                             )
@@ -1054,7 +1054,7 @@ To predict interactions/binding affinities of a single compound against a librar
                         with gr.Column():
                             HelpTip(
                                 "Select your preferred model, or click Recommend for the best-performing model based "
-                                "on the selected task, family, and whether the compound was trained."
+                                "on the selected task, family, and whether the compound was trained. "
                                 "Please refer to documentation for detailed benchamrk results."
                                 )
                             target_identify_preset = gr.Dropdown(list(PRESET_MAP.keys()), label='Step 5. Select a '
@@ -1080,15 +1080,23 @@ To predict interactions/binding affinities of a single compound against a librar
                                            visible=False)
         with gr.TabItem(label='Interaction pair inference', id=2):
             gr.Markdown('''
-# <center>DeepSEQreen Interaction Pair Inference</center>
+# <center>Interaction Pair Inference</center>
 <center>To predict interactions/binding affinities between any compound-protein pairs.</center>
 ''')
             with gr.Blocks() as infer_block:
                 with gr.Column() as infer_page:
+                    HelpTip(
+                        "A custom interation pair dataset can be a CSV file with 2 required columns "
+                        "(X1 for smiles and X2 for sequences) "
+                        "and optionally 2 ID columns (ID1 for compound ID and ID2 for target ID), "
+                        "or generated from a FASTA file containing multiple"
+                        "sequences and a SDF file containing multiple compounds."
+                    )
                     infer_type = gr.Dropdown(
-                        choices=['Upload a compound library and a target library',
-                                 'Upload a CSV interaction pair dataset'],
-                        value='Upload a compound library and a target library')
+                        choices=['Upload a CSV interaction pair dataset',
+                                 'Upload a compound library and a target library'],
+                        label='Step 1. Select Pair Input Type and Input',
+                        value='Upload a CSV interaction pair dataset')
                     with gr.Column() as pair_upload:
                         gr.File(label="Example custom dataset",
                                 value="data/examples/interaction_pair_inference.csv",
@@ -1113,18 +1121,39 @@ To predict interactions/binding affinities of a single compound against a librar
                             infer_target = gr.File(label='FASTA/CSV file containing multiple targets',
                                                    file_count="single", type='filepath')
 
-                    with gr.Row(visible=True):
-                        pair_infer_task = gr.Dropdown(list(TASK_MAP.keys()), label='Task')
-                        pair_infer_preset = gr.Dropdown(list(PRESET_MAP.keys()), label='Preset')
-                        pair_infer_target_family = gr.Dropdown(choices=['General'],
-                                                               label='Target family',
-                                                               value='General')
+                    with gr.Row():
+                        with gr.Column():
+                            HelpTip(
+                                "By default, models trained on all protein families (general) will be applied. "
+                                "If the proteins in the target library of interest all belong to the same protein family, manually selecting the family is supported."
+                            )
+                            pair_infer_target_family = gr.Dropdown(choices=list(TARGET_FAMILY_MAP.keys()),
+                                                                   value='General',
+                                                                   label='Step 2. Select Target Protein Family (Optional)')
 
-                    # with gr.Row():
-                    #     pair_infer_email = gr.Textbox(
-                    #         label='Email (optional)',
-                    #         info="Your email will be used to send you notifications when your job finishes."
-                    #     )
+                    with gr.Row():
+                        with gr.Column():
+                            HelpTip(
+                                "Interaction prediction provides you binding probability score between the target of interest and each compound in the library, "
+                                "while affinity prediction directly estimates their binding strength measured using IC50."
+                            )
+                            pair_infer_task = gr.Dropdown(list(TASK_MAP.keys()),
+                                                          label='Step 3. Select a Prediction Task',
+                                                          value='Compound-protein interaction')
+
+                    with gr.Row():
+                        with gr.Column():
+                            HelpTip("Select your preferred model. "
+                                    "Please refer to documentation for detailed benchamrk results."
+                                    )
+                            pair_infer_preset = gr.Dropdown(list(PRESET_MAP.keys()), label='Step 4. Select a Preset Model')
+                            # infer_preset_recommend_btn = gr.Button(value='Recommend a model', variant='primary')
+
+                    with gr.Row():
+                        pair_infer_email = gr.Textbox(
+                            label='Step 5. Email (Optional)',
+                            info="If an email is provided, a notification email will be sent to you when your job is completed."
+                        )
 
                     with gr.Row(visible=True):
                         # pair_infer_clr_btn = gr.ClearButton(size='lg')
@@ -1138,18 +1167,19 @@ To predict interactions/binding affinities of a single compound against a librar
         with gr.TabItem(label='Chemical property report', id=3):
             with gr.Blocks() as report:
                 gr.Markdown('''
-                # <center>DeepSEQreen Chemical Property Report</center>
-                <center>
+                # <center>Chemical Property Report</center>
+          
                 To compute chemical properties for the predictions of drug hit screening, 
-                target protein identification, and interaction pair inference. You may also upload 
-                your own dataset. 
+                target protein identification, and interaction pair inference. 
+                
+                You may also upload your own dataset using a CSV file containing one required column X1 for compound SMILES. 
                 
                 The page shows only a preview report displaying at most 30 records 
                 (with top predicted CPI/CPA if reporting results from a prediction job). 
                 
                 For a full report, please 
                 generate and download a CSV or interactive HTML report below.
-                </center>
+           
                 ''')
                 with gr.Row():
                     file_for_report = gr.File(interactive=True, type='filepath')
@@ -1160,7 +1190,7 @@ To predict interactions/binding affinities of a single compound against a librar
 
                 with gr.Row():
                     # clear_btn = gr.ClearButton(size='lg')
-                    analyze_btn = gr.Button('REPORT', variant='primary', size='lg', interactive=False)
+                    analyze_btn = gr.Button('Preview Top 30 Records', variant='primary', size='lg', interactive=False)
 
                 with gr.Row():
                     with gr.Column(scale=3):
@@ -1187,7 +1217,8 @@ To predict interactions/binding affinities of a single compound against a librar
                         gr.Textbox(visible=False, value=''),
                         gr.Textbox(visible=False, value=''),
                         gr.Button(visible=True),
-                        gr.Code(value='')]
+                        gr.Code(value=''),
+                        gr.Button(visible=False)]
             case 'Gene symbol':
                 return [gr.Dropdown(info=''),
                         gr.UploadButton(visible=False),
@@ -1195,7 +1226,8 @@ To predict interactions/binding affinities of a single compound against a librar
                         gr.Textbox(visible=True, value=''),
                         gr.Textbox(visible=True, value=''),
                         gr.Button(visible=True),
-                        gr.Code(value='')]
+                        gr.Code(value=''),
+                        gr.Button(visible=False)]
             case 'Sequence':
                 return [gr.Dropdown(info='Enter (paste) a FASTA string below manually or upload a FASTA file.'),
                         gr.UploadButton(visible=True),
@@ -1203,7 +1235,8 @@ To predict interactions/binding affinities of a single compound against a librar
                         gr.Textbox(visible=False, value=''),
                         gr.Textbox(visible=False, value=''),
                         gr.Button(visible=False),
-                        gr.Code(value='')]
+                        gr.Code(value=''),
+                        gr.Button(visible=True)]
 
 
     target_input_type.select(
@@ -1212,7 +1245,7 @@ To predict interactions/binding affinities of a single compound against a librar
         outputs=[
             target_input_type, target_upload_btn,
             target_id, target_gene, target_organism, target_query_btn,
-            target_fasta
+            target_fasta, target_paste_markdown
         ],
         show_progress=False
     )