import gradio as gr import sys sys.path.append('/home/user/app/ProteinMPNN-main/') from helper_scripts.parse_multiple_chains import * from protein_mpnn_run import * from helper_scripts.assign_fixed_chains import * from helper_scripts.make_fixed_positions_dict import * from helper_scripts.make_tied_positions_dict import * from helper_scripts.make_bias_AA import * sys.path.append('/home/user/app/file/') from Sort_Scores import * from Preinput_Merge import * from Sort_Dlkcat import * from Merge_Dlsc import * from Sort_Sco_Kcat import * sys.path.append('/home/user/app/DLKcat/DeeplearningApproach/Code/example/') # from model import * from prediction_for_input import * with gr.Blocks(css=".gradio-container {background-image: url('file=background.jpeg')}") as demo: gr.Markdown("Welcome using this demo.") with gr.Tab("HelloWorld"): gr.Markdown("Welcome using this demo.") gr.Markdown("This is a succend test") gr.Markdown("I think this demo can do some things") gr.Markdown("这一平台主要分为三部分，ProteinMPNN，DLKcat以及对文件的处理，可以在Files and versions 下看各自的README.md文件") with gr.Tab("helper_scripts"): with gr.Row(): input_files = gr.Files(label="input_files") input_file = gr.File(label="input_file") with gr.Column(): output_file = gr.Textbox(label="output_file") ca_only = gr.Checkbox(label="ca_only", value=False) chain_list = gr.Textbox(label="chain_list", value="") AA_list = gr.Textbox(label="AA_list", value="") with gr.Column(): position_list = gr.Textbox(label="position_list", value="") specify_non_fixed = gr.Checkbox(label="specify_non_fixed", value=False) homooligomer = gr.Number(label="homooligomer", value=0) bias_list = gr.Textbox(label="bias_list", value="") output = gr.File() with gr.Row(): pmc_button = gr.Button("parse_multiple_chains") afc_button = gr.Button("assign_fixed_chains") mfpd_button = gr.Button("make_fixed_positions_dict") mtpd_button = gr.Button("make_tied_positions_dict") mbA_button = gr.Button("make_bias_AA") pmc_button.click(p_m_c, inputs=[input_files, output_file, ca_only], outputs=output) afc_button.click(a_f_c, inputs=[input_file, output_file, chain_list], outputs=output) mfpd_button.click(m_f_p_d, inputs=[input_file, output_file, chain_list, position_list, specify_non_fixed], outputs=output) mtpd_button.click(m_t_p_d, inputs=[input_file, output_file, chain_list, position_list, homooligomer], outputs=output) mbA_button.click(m_b_A, inputs=[output_file, AA_list, bias_list], outputs=output) with gr.Tab("protein_mpnn_run"): with gr.Box(): ca_only = gr.Checkbox(label="ca_only", value=False, visible=False) path_to_model_weights = gr.Textbox(label="path_to_model_weights", value="", visible=False) model_name = gr.Textbox(label="model_name", value="v_48_020", visible=False) with gr.Row(): seed = gr.Number(value=int(0), label="seed") save_score = gr.Number(label="save_score", value=int(0), visible=False) save_probs = gr.Number(label="save_probs", value=int(0), visible=False) score_only = gr.Number(label="score_only", value=int(0)) conditional_probs_only = gr.Number(label="conditional_probs_only", value=int(0), visible=False) conditional_probs_only_backbone = gr.Number(label="conditional_probs_only_backbone", value=int(0), visible=False) with gr.Row(): unconditional_probs_only = gr.Number(label="unconditional_probs_only", value=int(0)) backbone_noise = gr.Number(label="backbone_noise", value=0.00, visible=False) num_seq_per_target = gr.Number(value=int(1), label="num_seq_per_target") with gr.Row(): batch_size = gr.Number(value=int(1), label="batch_size") max_length = gr.Number(label="max_length", value=int(200000), visible=False) sampling_temp = gr.Textbox(value="0.1", label="sampling_temp") out_folder = gr.Textbox(label="out_folder") pdb_path_chains = gr.Textbox(label="pdb_path_chains", value="") with gr.Box(): with gr.Row(): pdb_path = gr.File(label="pdb_path") jsonl_path = gr.File(label="jsonl_path") with gr.Row(): chain_id_jsonl = gr.File(label="chain_id_jsonl") fixed_positions_jsonl = gr.File(label="fixed_positions_jsonl") omit_AAs = gr.Textbox(label="omit_AAs", value="X", visible=False) with gr.Row(): bias_AA_jsonl = gr.File(label="bias_AA_jsonl") tied_positions_jsonl = gr.File(label="tied_positions_jsonl") bias_by_res_jsonl = gr.Textbox(label="bias_by_res_jsonl", value="", visible=False) omit_AA_jsonl = gr.Textbox(label="omit_AA_jsonl", value="", visible=False) pssm_jsonl = gr.Textbox(label="pssm_jsonl", value="", visible=False) pssm_multi = gr.Number(label="pssm_multi", value=0.0, visible=False) pssm_threshold = gr.Number(label="pssm_threshold", value=0.0, visible=False) pssm_log_odds_flag = gr.Number(label="pssm_log_odds_flag", value=int(0), visible=False) pssm_bias_flag = gr.Number(label="pssm_bias_flag", value=int(0), visible=False) pmroutput = gr.Files() pmr_button = gr.Button("protein_mpnn_run") pmr_button.click(p_m_r, inputs=[ca_only, path_to_model_weights, model_name, seed, save_score, save_probs, score_only, conditional_probs_only, conditional_probs_only_backbone , unconditional_probs_only, backbone_noise, num_seq_per_target, batch_size, max_length, sampling_temp, out_folder, pdb_path, pdb_path_chains, jsonl_path, chain_id_jsonl, fixed_positions_jsonl, omit_AAs, bias_AA_jsonl, bias_by_res_jsonl , omit_AA_jsonl, pssm_jsonl, pssm_multi, pssm_threshold, pssm_log_odds_flag, pssm_bias_flag, tied_positions_jsonl], outputs=pmroutput) with gr.Tab("Sort"): file1_input = gr.File(label="输入相关文件") file1_output = gr.File() with gr.Row(): file1_button1 = gr.Button("Sort Scores") file1_button2 = gr.Button("Sort Dlkcat") file1_button3 = gr.Button("Sort Mergekcat") with gr.Tab("Pre Merge"): with gr.Row(): file2_input1 = gr.File(label="strip_file") file2_input2 = gr.File(label="smi_file") file2_input3 = gr.File(label="seq_file") file2_output = gr.File() file2_button1 = gr.Button("Strip") with gr.Row(): file2_button2 = gr.Button("Merge") file2_button3 = gr.Button("Merge All") with gr.Tab("Merge Dlsc"): with gr.Row(): file4_input1 = gr.File(label="sc_file") file4_input2 = gr.File(label="cat_file") file4_output = gr.File() file4_button = gr.Button("Merge") with gr.Tab("Prediction Dlkcat"): file6_input = gr.File(label="输入相关文件") file6_output = gr.File() file6_button = gr.Button("Prediction_Dlkcat") with gr.Accordion("Open for More!"): gr.Markdown("Look at me...") file1_button1.click(Sort_Scores, inputs=file1_input, outputs=file1_output) file1_button2.click(Sort_Dlkcat, inputs=file1_input, outputs=file1_output) file1_button3.click(Sort_Sco_Kcat, inputs=file1_input, outputs=file1_output) file2_button1.click(Strip, inputs=file2_input1, outputs=file2_output) file2_button2.click(Merge, inputs=[file2_input2, file2_input3], outputs=file2_output) file2_button3.click(Merge_All, inputs=[file2_input2, file2_input3], outputs=file2_output) file4_button.click(Merge_Dlsc, inputs=[file4_input1, file4_input2], outputs=file4_output) file6_button.click(test, inputs=file6_input, outputs=file6_output) if __name__ == "__main__": demo.launch()