Spaces:
Paused
Paused
Delete app-backup.py
Browse files- app-backup.py +0 -591
app-backup.py
DELETED
|
@@ -1,591 +0,0 @@
|
|
| 1 |
-
import os,sys
|
| 2 |
-
|
| 3 |
-
# install environment goods
|
| 4 |
-
#os.system("pip -q install dgl -f https://data.dgl.ai/wheels/cu113/repo.html")
|
| 5 |
-
os.system('pip install dgl==1.0.2+cu116 -f https://data.dgl.ai/wheels/cu116/repo.html')
|
| 6 |
-
#os.system('pip install gradio')
|
| 7 |
-
os.environ["DGLBACKEND"] = "pytorch"
|
| 8 |
-
#os.system(f'pip install -r ./PROTEIN_GENERATOR/requirements.txt')
|
| 9 |
-
print('Modules installed')
|
| 10 |
-
|
| 11 |
-
#os.system('pip install --force gradio==3.36.1')
|
| 12 |
-
#os.system('pip install gradio_client==0.2.7')
|
| 13 |
-
#os.system('pip install \"numpy<2\"')
|
| 14 |
-
#os.system('pip install numpy --upgrade')
|
| 15 |
-
#os.system('pip install --force numpy==1.24.1')
|
| 16 |
-
|
| 17 |
-
|
| 18 |
-
if not os.path.exists('./SEQDIFF_230205_dssp_hotspots_25mask_EQtasks_mod30.pt'):
|
| 19 |
-
print('Downloading model weights 1')
|
| 20 |
-
os.system('wget http://files.ipd.uw.edu/pub/sequence_diffusion/checkpoints/SEQDIFF_230205_dssp_hotspots_25mask_EQtasks_mod30.pt')
|
| 21 |
-
print('Successfully Downloaded')
|
| 22 |
-
|
| 23 |
-
if not os.path.exists('./SEQDIFF_221219_equalTASKS_nostrSELFCOND_mod30.pt'):
|
| 24 |
-
print('Downloading model weights 2')
|
| 25 |
-
os.system('wget http://files.ipd.uw.edu/pub/sequence_diffusion/checkpoints/SEQDIFF_221219_equalTASKS_nostrSELFCOND_mod30.pt')
|
| 26 |
-
print('Successfully Downloaded')
|
| 27 |
-
|
| 28 |
-
import numpy as np
|
| 29 |
-
import gradio as gr
|
| 30 |
-
import py3Dmol
|
| 31 |
-
from io import StringIO
|
| 32 |
-
import json
|
| 33 |
-
import secrets
|
| 34 |
-
import copy
|
| 35 |
-
import matplotlib.pyplot as plt
|
| 36 |
-
from utils.sampler import HuggingFace_sampler
|
| 37 |
-
from utils.parsers_inference import parse_pdb
|
| 38 |
-
from model.util import writepdb
|
| 39 |
-
from utils.inpainting_util import *
|
| 40 |
-
|
| 41 |
-
|
| 42 |
-
plt.rcParams.update({'font.size': 13})
|
| 43 |
-
|
| 44 |
-
with open('./tmp/args.json','r') as f:
|
| 45 |
-
args = json.load(f)
|
| 46 |
-
|
| 47 |
-
# manually set checkpoint to load
|
| 48 |
-
args['checkpoint'] = None
|
| 49 |
-
args['dump_trb'] = False
|
| 50 |
-
args['dump_args'] = True
|
| 51 |
-
args['save_best_plddt'] = True
|
| 52 |
-
args['T'] = 25
|
| 53 |
-
args['strand_bias'] = 0.0
|
| 54 |
-
args['loop_bias'] = 0.0
|
| 55 |
-
args['helix_bias'] = 0.0
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
| 59 |
-
def protein_diffusion_model(sequence, seq_len, helix_bias, strand_bias, loop_bias,
|
| 60 |
-
secondary_structure, aa_bias, aa_bias_potential,
|
| 61 |
-
#target_charge, target_ph, charge_potential,
|
| 62 |
-
num_steps, noise, hydrophobic_target_score, hydrophobic_potential,
|
| 63 |
-
contigs, pssm, seq_mask, str_mask, rewrite_pdb):
|
| 64 |
-
|
| 65 |
-
dssp_checkpoint = './SEQDIFF_230205_dssp_hotspots_25mask_EQtasks_mod30.pt'
|
| 66 |
-
og_checkpoint = './SEQDIFF_221219_equalTASKS_nostrSELFCOND_mod30.pt'
|
| 67 |
-
|
| 68 |
-
model_args = copy.deepcopy(args)
|
| 69 |
-
|
| 70 |
-
# make sampler
|
| 71 |
-
S = HuggingFace_sampler(args=model_args)
|
| 72 |
-
|
| 73 |
-
# get random prefix
|
| 74 |
-
S.out_prefix = './tmp/'+secrets.token_hex(nbytes=10).upper()
|
| 75 |
-
|
| 76 |
-
# set args
|
| 77 |
-
S.args['checkpoint'] = None
|
| 78 |
-
S.args['dump_trb'] = False
|
| 79 |
-
S.args['dump_args'] = True
|
| 80 |
-
S.args['save_best_plddt'] = True
|
| 81 |
-
S.args['T'] = 20
|
| 82 |
-
S.args['strand_bias'] = 0.0
|
| 83 |
-
S.args['loop_bias'] = 0.0
|
| 84 |
-
S.args['helix_bias'] = 0.0
|
| 85 |
-
S.args['potentials'] = None
|
| 86 |
-
S.args['potential_scale'] = None
|
| 87 |
-
S.args['aa_composition'] = None
|
| 88 |
-
|
| 89 |
-
|
| 90 |
-
# get sequence if entered and make sure all chars are valid
|
| 91 |
-
alt_aa_dict = {'B':['D','N'],'J':['I','L'],'U':['C'],'Z':['E','Q'],'O':['K']}
|
| 92 |
-
if sequence not in ['',None]:
|
| 93 |
-
L = len(sequence)
|
| 94 |
-
aa_seq = []
|
| 95 |
-
for aa in sequence.upper():
|
| 96 |
-
if aa in alt_aa_dict.keys():
|
| 97 |
-
aa_seq.append(np.random.choice(alt_aa_dict[aa]))
|
| 98 |
-
else:
|
| 99 |
-
aa_seq.append(aa)
|
| 100 |
-
|
| 101 |
-
S.args['sequence'] = aa_seq
|
| 102 |
-
elif contigs not in ['',None]:
|
| 103 |
-
S.args['contigs'] = [contigs]
|
| 104 |
-
else:
|
| 105 |
-
S.args['contigs'] = [f'{seq_len}']
|
| 106 |
-
L = int(seq_len)
|
| 107 |
-
|
| 108 |
-
print('DEBUG: ',rewrite_pdb)
|
| 109 |
-
if rewrite_pdb not in ['',None]:
|
| 110 |
-
S.args['pdb'] = rewrite_pdb.name
|
| 111 |
-
|
| 112 |
-
if seq_mask not in ['',None]:
|
| 113 |
-
S.args['inpaint_seq'] = [seq_mask]
|
| 114 |
-
if str_mask not in ['',None]:
|
| 115 |
-
S.args['inpaint_str'] = [str_mask]
|
| 116 |
-
|
| 117 |
-
if secondary_structure in ['',None]:
|
| 118 |
-
secondary_structure = None
|
| 119 |
-
else:
|
| 120 |
-
secondary_structure = ''.join(['E' if x == 'S' else x for x in secondary_structure])
|
| 121 |
-
if L < len(secondary_structure):
|
| 122 |
-
secondary_structure = secondary_structure[:len(sequence)]
|
| 123 |
-
elif L == len(secondary_structure):
|
| 124 |
-
pass
|
| 125 |
-
else:
|
| 126 |
-
dseq = L - len(secondary_structure)
|
| 127 |
-
secondary_structure += secondary_structure[-1]*dseq
|
| 128 |
-
|
| 129 |
-
|
| 130 |
-
# potentials
|
| 131 |
-
potential_list = []
|
| 132 |
-
potential_bias_list = []
|
| 133 |
-
|
| 134 |
-
if aa_bias not in ['',None]:
|
| 135 |
-
potential_list.append('aa_bias')
|
| 136 |
-
S.args['aa_composition'] = aa_bias
|
| 137 |
-
if aa_bias_potential in ['',None]:
|
| 138 |
-
aa_bias_potential = 3
|
| 139 |
-
potential_bias_list.append(str(aa_bias_potential))
|
| 140 |
-
'''
|
| 141 |
-
if target_charge not in ['',None]:
|
| 142 |
-
potential_list.append('charge')
|
| 143 |
-
if charge_potential in ['',None]:
|
| 144 |
-
charge_potential = 1
|
| 145 |
-
potential_bias_list.append(str(charge_potential))
|
| 146 |
-
S.args['target_charge'] = float(target_charge)
|
| 147 |
-
if target_ph in ['',None]:
|
| 148 |
-
target_ph = 7.4
|
| 149 |
-
S.args['target_pH'] = float(target_ph)
|
| 150 |
-
'''
|
| 151 |
-
|
| 152 |
-
if hydrophobic_target_score not in ['',None]:
|
| 153 |
-
potential_list.append('hydrophobic')
|
| 154 |
-
S.args['hydrophobic_score'] = float(hydrophobic_target_score)
|
| 155 |
-
if hydrophobic_potential in ['',None]:
|
| 156 |
-
hydrophobic_potential = 3
|
| 157 |
-
potential_bias_list.append(str(hydrophobic_potential))
|
| 158 |
-
|
| 159 |
-
if pssm not in ['',None]:
|
| 160 |
-
potential_list.append('PSSM')
|
| 161 |
-
potential_bias_list.append('5')
|
| 162 |
-
S.args['PSSM'] = pssm.name
|
| 163 |
-
|
| 164 |
-
|
| 165 |
-
if len(potential_list) > 0:
|
| 166 |
-
S.args['potentials'] = ','.join(potential_list)
|
| 167 |
-
S.args['potential_scale'] = ','.join(potential_bias_list)
|
| 168 |
-
|
| 169 |
-
|
| 170 |
-
# normalise secondary_structure bias from range 0-0.3
|
| 171 |
-
S.args['secondary_structure'] = secondary_structure
|
| 172 |
-
S.args['helix_bias'] = helix_bias
|
| 173 |
-
S.args['strand_bias'] = strand_bias
|
| 174 |
-
S.args['loop_bias'] = loop_bias
|
| 175 |
-
|
| 176 |
-
# set T
|
| 177 |
-
if num_steps in ['',None]:
|
| 178 |
-
S.args['T'] = 20
|
| 179 |
-
else:
|
| 180 |
-
S.args['T'] = int(num_steps)
|
| 181 |
-
|
| 182 |
-
# noise
|
| 183 |
-
if 'normal' in noise:
|
| 184 |
-
S.args['sample_distribution'] = noise
|
| 185 |
-
S.args['sample_distribution_gmm_means'] = [0]
|
| 186 |
-
S.args['sample_distribution_gmm_variances'] = [1]
|
| 187 |
-
elif 'gmm2' in noise:
|
| 188 |
-
S.args['sample_distribution'] = noise
|
| 189 |
-
S.args['sample_distribution_gmm_means'] = [-1,1]
|
| 190 |
-
S.args['sample_distribution_gmm_variances'] = [1,1]
|
| 191 |
-
elif 'gmm3' in noise:
|
| 192 |
-
S.args['sample_distribution'] = noise
|
| 193 |
-
S.args['sample_distribution_gmm_means'] = [-1,0,1]
|
| 194 |
-
S.args['sample_distribution_gmm_variances'] = [1,1,1]
|
| 195 |
-
|
| 196 |
-
|
| 197 |
-
|
| 198 |
-
if secondary_structure not in ['',None] or helix_bias+strand_bias+loop_bias > 0:
|
| 199 |
-
S.args['checkpoint'] = dssp_checkpoint
|
| 200 |
-
S.args['d_t1d'] = 29
|
| 201 |
-
print('using dssp checkpoint')
|
| 202 |
-
else:
|
| 203 |
-
S.args['checkpoint'] = og_checkpoint
|
| 204 |
-
S.args['d_t1d'] = 24
|
| 205 |
-
print('using og checkpoint')
|
| 206 |
-
|
| 207 |
-
|
| 208 |
-
for k,v in S.args.items():
|
| 209 |
-
print(f"{k} --> {v}")
|
| 210 |
-
|
| 211 |
-
# init S
|
| 212 |
-
S.model_init()
|
| 213 |
-
S.diffuser_init()
|
| 214 |
-
S.setup()
|
| 215 |
-
|
| 216 |
-
# sampling loop
|
| 217 |
-
plddt_data = []
|
| 218 |
-
for j in range(S.max_t):
|
| 219 |
-
print(f'on step {j}')
|
| 220 |
-
output_seq, output_pdb, plddt = S.take_step_get_outputs(j)
|
| 221 |
-
plddt_data.append(plddt)
|
| 222 |
-
yield output_seq, output_pdb, display_pdb(output_pdb), get_plddt_plot(plddt_data, S.max_t)
|
| 223 |
-
|
| 224 |
-
output_seq, output_pdb, plddt = S.get_outputs()
|
| 225 |
-
yield output_seq, output_pdb, display_pdb(output_pdb), get_plddt_plot(plddt_data, S.max_t)
|
| 226 |
-
|
| 227 |
-
def get_plddt_plot(plddt_data, max_t):
|
| 228 |
-
x = [i+1 for i in range(len(plddt_data))]
|
| 229 |
-
fig, ax = plt.subplots(figsize=(15,6))
|
| 230 |
-
ax.plot(x,plddt_data,color='#661dbf', linewidth=3,marker='o')
|
| 231 |
-
ax.set_xticks([i+1 for i in range(max_t)])
|
| 232 |
-
ax.set_yticks([(i+1)/10 for i in range(10)])
|
| 233 |
-
ax.set_ylim([0,1])
|
| 234 |
-
ax.set_ylabel('model confidence (plddt)')
|
| 235 |
-
ax.set_xlabel('diffusion steps (t)')
|
| 236 |
-
return fig
|
| 237 |
-
|
| 238 |
-
def display_pdb(path_to_pdb):
|
| 239 |
-
'''
|
| 240 |
-
#function to display pdb in py3dmol
|
| 241 |
-
'''
|
| 242 |
-
pdb = open(path_to_pdb, "r").read()
|
| 243 |
-
|
| 244 |
-
view = py3Dmol.view(width=500, height=500)
|
| 245 |
-
view.addModel(pdb, "pdb")
|
| 246 |
-
view.setStyle({'model': -1}, {"cartoon": {'colorscheme':{'prop':'b','gradient':'roygb','min':0,'max':1}}})#'linear', 'min': 0, 'max': 1, 'colors': ["#ff9ef0","#a903fc",]}}})
|
| 247 |
-
view.zoomTo()
|
| 248 |
-
output = view._make_html().replace("'", '"')
|
| 249 |
-
print(view._make_html())
|
| 250 |
-
x = f"""<!DOCTYPE html><html></center> {output} </center></html>""" # do not use ' in this input
|
| 251 |
-
|
| 252 |
-
return f"""<iframe height="500px" width="100%" name="result" allow="midi; geolocation; microphone; camera;
|
| 253 |
-
display-capture; encrypted-media;" sandbox="allow-modals allow-forms
|
| 254 |
-
allow-scripts allow-same-origin allow-popups
|
| 255 |
-
allow-top-navigation-by-user-activation allow-downloads" allowfullscreen=""
|
| 256 |
-
allowpaymentrequest="" frameborder="0" srcdoc='{x}'></iframe>"""
|
| 257 |
-
|
| 258 |
-
'''
|
| 259 |
-
return f"""<iframe style="width: 100%; height:700px" name="result" allow="midi; geolocation; microphone; camera;
|
| 260 |
-
display-capture; encrypted-media;" sandbox="allow-modals allow-forms
|
| 261 |
-
allow-scripts allow-same-origin allow-popups
|
| 262 |
-
allow-top-navigation-by-user-activation allow-downloads" allowfullscreen=""
|
| 263 |
-
allowpaymentrequest="" frameborder="0" srcdoc='{x}'></iframe>"""
|
| 264 |
-
'''
|
| 265 |
-
|
| 266 |
-
|
| 267 |
-
|
| 268 |
-
# MOTIF SCAFFOLDING
|
| 269 |
-
def get_motif_preview(pdb_id, contigs):
|
| 270 |
-
'''
|
| 271 |
-
#function to display selected motif in py3dmol
|
| 272 |
-
'''
|
| 273 |
-
input_pdb = fetch_pdb(pdb_id=pdb_id.lower())
|
| 274 |
-
|
| 275 |
-
# rewrite pdb
|
| 276 |
-
parse = parse_pdb(input_pdb)
|
| 277 |
-
#output_name = './rewrite_'+input_pdb.split('/')[-1]
|
| 278 |
-
#writepdb(output_name, torch.tensor(parse_og['xyz']),torch.tensor(parse_og['seq']))
|
| 279 |
-
#parse = parse_pdb(output_name)
|
| 280 |
-
output_name = input_pdb
|
| 281 |
-
|
| 282 |
-
pdb = open(output_name, "r").read()
|
| 283 |
-
view = py3Dmol.view(width=500, height=500)
|
| 284 |
-
view.addModel(pdb, "pdb")
|
| 285 |
-
|
| 286 |
-
if contigs in ['',0]:
|
| 287 |
-
contigs = ['0']
|
| 288 |
-
else:
|
| 289 |
-
contigs = [contigs]
|
| 290 |
-
|
| 291 |
-
print('DEBUG: ',contigs)
|
| 292 |
-
|
| 293 |
-
pdb_map = get_mappings(ContigMap(parse,contigs))
|
| 294 |
-
print('DEBUG: ',pdb_map)
|
| 295 |
-
print('DEBUG: ',pdb_map['con_ref_idx0'])
|
| 296 |
-
roi = [x[1]-1 for x in pdb_map['con_ref_pdb_idx']]
|
| 297 |
-
|
| 298 |
-
colormap = {0:'#D3D3D3', 1:'#F74CFF'}
|
| 299 |
-
colors = {i+1: colormap[1] if i in roi else colormap[0] for i in range(parse['xyz'].shape[0])}
|
| 300 |
-
view.setStyle({"cartoon": {"colorscheme": {"prop": "resi", "map": colors}}})
|
| 301 |
-
view.zoomTo()
|
| 302 |
-
output = view._make_html().replace("'", '"')
|
| 303 |
-
print(view._make_html())
|
| 304 |
-
x = f"""<!DOCTYPE html><html></center> {output} </center></html>""" # do not use ' in this input
|
| 305 |
-
|
| 306 |
-
return f"""<iframe height="500px" width="100%" name="result" allow="midi; geolocation; microphone; camera;
|
| 307 |
-
display-capture; encrypted-media;" sandbox="allow-modals allow-forms
|
| 308 |
-
allow-scripts allow-same-origin allow-popups
|
| 309 |
-
allow-top-navigation-by-user-activation allow-downloads" allowfullscreen=""
|
| 310 |
-
allowpaymentrequest="" frameborder="0" srcdoc='{x}'></iframe>""", output_name
|
| 311 |
-
|
| 312 |
-
def fetch_pdb(pdb_id=None):
|
| 313 |
-
if pdb_id is None or pdb_id == "":
|
| 314 |
-
return None
|
| 315 |
-
else:
|
| 316 |
-
os.system(f"wget -qnc https://files.rcsb.org/view/{pdb_id}.pdb")
|
| 317 |
-
return f"{pdb_id}.pdb"
|
| 318 |
-
|
| 319 |
-
# MSA AND PSSM GUIDANCE
|
| 320 |
-
def save_pssm(file_upload):
|
| 321 |
-
filename = file_upload.name
|
| 322 |
-
orig_name = file_upload.orig_name
|
| 323 |
-
if filename.split('.')[-1] in ['fasta', 'a3m']:
|
| 324 |
-
return msa_to_pssm(file_upload)
|
| 325 |
-
return filename
|
| 326 |
-
|
| 327 |
-
def msa_to_pssm(msa_file):
|
| 328 |
-
# Define the lookup table for converting amino acids to indices
|
| 329 |
-
aa_to_index = {'A': 0, 'R': 1, 'N': 2, 'D': 3, 'C': 4, 'Q': 5, 'E': 6, 'G': 7, 'H': 8, 'I': 9, 'L': 10,
|
| 330 |
-
'K': 11, 'M': 12, 'F': 13, 'P': 14, 'S': 15, 'T': 16, 'W': 17, 'Y': 18, 'V': 19, 'X': 20, '-': 21}
|
| 331 |
-
# Open the FASTA file and read the sequences
|
| 332 |
-
records = list(SeqIO.parse(msa_file.name, "fasta"))
|
| 333 |
-
|
| 334 |
-
assert len(records) >= 1, "MSA must contain more than one protein sequecne."
|
| 335 |
-
|
| 336 |
-
first_seq = str(records[0].seq)
|
| 337 |
-
aligned_seqs = [first_seq]
|
| 338 |
-
# print(aligned_seqs)
|
| 339 |
-
# Perform sequence alignment using the Needleman-Wunsch algorithm
|
| 340 |
-
aligner = Align.PairwiseAligner()
|
| 341 |
-
aligner.open_gap_score = -0.7
|
| 342 |
-
aligner.extend_gap_score = -0.3
|
| 343 |
-
for record in records[1:]:
|
| 344 |
-
alignment = aligner.align(first_seq, str(record.seq))[0]
|
| 345 |
-
alignment = alignment.format().split("\n")
|
| 346 |
-
al1 = alignment[0]
|
| 347 |
-
al2 = alignment[2]
|
| 348 |
-
al1_fin = ""
|
| 349 |
-
al2_fin = ""
|
| 350 |
-
percent_gap = al2.count('-')/ len(al2)
|
| 351 |
-
if percent_gap > 0.4:
|
| 352 |
-
continue
|
| 353 |
-
for i in range(len(al1)):
|
| 354 |
-
if al1[i] != '-':
|
| 355 |
-
al1_fin += al1[i]
|
| 356 |
-
al2_fin += al2[i]
|
| 357 |
-
aligned_seqs.append(str(al2_fin))
|
| 358 |
-
# Get the length of the aligned sequences
|
| 359 |
-
aligned_seq_length = len(first_seq)
|
| 360 |
-
# Initialize the position scoring matrix
|
| 361 |
-
matrix = np.zeros((22, aligned_seq_length))
|
| 362 |
-
# Iterate through the aligned sequences and count the amino acids at each position
|
| 363 |
-
for seq in aligned_seqs:
|
| 364 |
-
#print(seq)
|
| 365 |
-
for i in range(aligned_seq_length):
|
| 366 |
-
if i == len(seq):
|
| 367 |
-
break
|
| 368 |
-
amino_acid = seq[i]
|
| 369 |
-
if amino_acid.upper() not in aa_to_index.keys():
|
| 370 |
-
continue
|
| 371 |
-
else:
|
| 372 |
-
aa_index = aa_to_index[amino_acid.upper()]
|
| 373 |
-
matrix[aa_index, i] += 1
|
| 374 |
-
# Normalize the counts to get the frequency of each amino acid at each position
|
| 375 |
-
matrix /= len(aligned_seqs)
|
| 376 |
-
print(len(aligned_seqs))
|
| 377 |
-
matrix[20:,]=0
|
| 378 |
-
|
| 379 |
-
outdir = ".".join(msa_file.name.split('.')[:-1]) + ".csv"
|
| 380 |
-
np.savetxt(outdir, matrix[:21,:].T, delimiter=",")
|
| 381 |
-
return outdir
|
| 382 |
-
|
| 383 |
-
def get_pssm(fasta_msa, input_pssm):
|
| 384 |
-
|
| 385 |
-
if input_pssm not in ['',None]:
|
| 386 |
-
outdir = input_pssm.name
|
| 387 |
-
else:
|
| 388 |
-
outdir = save_pssm(fasta_msa)
|
| 389 |
-
|
| 390 |
-
pssm = np.loadtxt(outdir, delimiter=",", dtype=float)
|
| 391 |
-
fig, ax = plt.subplots(figsize=(15,6))
|
| 392 |
-
plt.imshow(torch.permute(torch.tensor(pssm),(1,0)))
|
| 393 |
-
|
| 394 |
-
return fig, outdir
|
| 395 |
-
|
| 396 |
-
|
| 397 |
-
#toggle options
|
| 398 |
-
def toggle_seq_input(choice):
|
| 399 |
-
if choice == "protein length":
|
| 400 |
-
return gr.update(visible=True, value=None), gr.update(visible=False, value=None)
|
| 401 |
-
elif choice == "custom sequence":
|
| 402 |
-
return gr.update(visible=False, value=None), gr.update(visible=True, value=None)
|
| 403 |
-
|
| 404 |
-
def toggle_secondary_structure(choice):
|
| 405 |
-
if choice == "sliders":
|
| 406 |
-
return gr.update(visible=True, value=None),gr.update(visible=True, value=None),gr.update(visible=True, value=None),gr.update(visible=False, value=None)
|
| 407 |
-
elif choice == "explicit":
|
| 408 |
-
return gr.update(visible=False, value=None),gr.update(visible=False, value=None),gr.update(visible=False, value=None),gr.update(visible=True, value=None)
|
| 409 |
-
|
| 410 |
-
|
| 411 |
-
# Define the Gradio interface
|
| 412 |
-
with gr.Blocks(theme='ParityError/Interstellar') as demo:
|
| 413 |
-
|
| 414 |
-
|
| 415 |
-
|
| 416 |
-
#with gr.Row().style(equal_height=False):
|
| 417 |
-
with gr.Row():
|
| 418 |
-
with gr.Column():
|
| 419 |
-
with gr.Tabs():
|
| 420 |
-
with gr.TabItem("Inputs"):
|
| 421 |
-
gr.Markdown("""## INPUTS""")
|
| 422 |
-
gr.Markdown("""#### Start Sequence
|
| 423 |
-
Specify the protein length for complete unconditional generation, or scaffold a motif (or your name) using the custom sequence input""")
|
| 424 |
-
seq_opt = gr.Radio(["protein length","custom sequence"], label="How would you like to specify the starting sequence?", value='protein length')
|
| 425 |
-
|
| 426 |
-
sequence = gr.Textbox(label="custom sequence", lines=1, placeholder='AMINO ACIDS: A,C,D,E,F,G,H,I,K,L,M,N,P,Q,R,S,T,V,W,Y\n MASK TOKEN: X', visible=False)
|
| 427 |
-
seq_len = gr.Slider(minimum=5.0, maximum=250.0, label="protein length", value=100, visible=True)
|
| 428 |
-
|
| 429 |
-
seq_opt.change(fn=toggle_seq_input,
|
| 430 |
-
inputs=[seq_opt],
|
| 431 |
-
outputs=[seq_len, sequence],
|
| 432 |
-
queue=False)
|
| 433 |
-
|
| 434 |
-
gr.Markdown("""### Optional Parameters""")
|
| 435 |
-
with gr.Accordion(label='Secondary Structure',open=True):
|
| 436 |
-
gr.Markdown("""Try changing the sliders or inputing explicit secondary structure conditioning for each residue""")
|
| 437 |
-
sec_str_opt = gr.Radio(["sliders","explicit"], label="How would you like to specify secondary structure?", value='sliders')
|
| 438 |
-
|
| 439 |
-
secondary_structure = gr.Textbox(label="secondary structure", lines=1, placeholder='HELIX = H STRAND = S LOOP = L MASK = X(must be the same length as input sequence)', visible=False)
|
| 440 |
-
|
| 441 |
-
with gr.Column():
|
| 442 |
-
helix_bias = gr.Slider(minimum=0.0, maximum=0.05, label="helix bias", visible=True)
|
| 443 |
-
strand_bias = gr.Slider(minimum=0.0, maximum=0.05, label="strand bias", visible=True)
|
| 444 |
-
loop_bias = gr.Slider(minimum=0.0, maximum=0.20, label="loop bias", visible=True)
|
| 445 |
-
|
| 446 |
-
sec_str_opt.change(fn=toggle_secondary_structure,
|
| 447 |
-
inputs=[sec_str_opt],
|
| 448 |
-
outputs=[helix_bias,strand_bias,loop_bias,secondary_structure],
|
| 449 |
-
queue=False)
|
| 450 |
-
|
| 451 |
-
with gr.Accordion(label='Amino Acid Compositional Bias',open=False):
|
| 452 |
-
gr.Markdown("""Bias sequence composition for particular amino acids by specifying the one letter code followed by the fraction to bias. This can be input as a list for example: W0.2,E0.1""")
|
| 453 |
-
with gr.Row():
|
| 454 |
-
aa_bias = gr.Textbox(label="aa bias", lines=1, placeholder='specify one letter AA and fraction to bias, for example W0.1 or M0.1,K0.1' )
|
| 455 |
-
aa_bias_potential = gr.Textbox(label="aa bias scale", lines=1, placeholder='AA Bias potential scale (recomended range 1.0-5.0)')
|
| 456 |
-
|
| 457 |
-
'''
|
| 458 |
-
with gr.Accordion(label='Charge Bias',open=False):
|
| 459 |
-
gr.Markdown("""Bias for a specified net charge at a particular pH using the boxes below""")
|
| 460 |
-
with gr.Row():
|
| 461 |
-
target_charge = gr.Textbox(label="net charge", lines=1, placeholder='net charge to target')
|
| 462 |
-
target_ph = gr.Textbox(label="pH", lines=1, placeholder='pH at which net charge is desired')
|
| 463 |
-
charge_potential = gr.Textbox(label="charge potential scale", lines=1, placeholder='charge potential scale (recomended range 1.0-5.0)')
|
| 464 |
-
'''
|
| 465 |
-
|
| 466 |
-
with gr.Accordion(label='Hydrophobic Bias',open=False):
|
| 467 |
-
gr.Markdown("""Bias for or against hydrophobic composition, to get more soluble proteins, bias away with a negative target score (ex. -5)""")
|
| 468 |
-
with gr.Row():
|
| 469 |
-
hydrophobic_target_score = gr.Textbox(label="hydrophobic score", lines=1, placeholder='hydrophobic score to target (negative score is good for solublility)')
|
| 470 |
-
hydrophobic_potential = gr.Textbox(label="hydrophobic potential scale", lines=1, placeholder='hydrophobic potential scale (recomended range 1.0-2.0)')
|
| 471 |
-
|
| 472 |
-
with gr.Accordion(label='Diffusion Params',open=False):
|
| 473 |
-
gr.Markdown("""Increasing T to more steps can be helpful for harder design challenges, sampling from different distributions can change the sequence and structural composition""")
|
| 474 |
-
with gr.Row():
|
| 475 |
-
num_steps = gr.Textbox(label="T", lines=1, placeholder='number of diffusion steps (25 or less will speed things up)')
|
| 476 |
-
noise = gr.Dropdown(['normal','gmm2 [-1,1]','gmm3 [-1,0,1]'], label='noise type', value='normal')
|
| 477 |
-
|
| 478 |
-
with gr.TabItem("Motif Selection"):
|
| 479 |
-
|
| 480 |
-
gr.Markdown("""### Motif Selection Preview""")
|
| 481 |
-
gr.Markdown('Contigs explained: to grab residues (seq and str) on a pdb chain you will provide the chain letter followed by a range of residues as indexed in the pdb file for example (A3-10) is the syntax to select residues 3-10 on chain A (the chain always needs to be specified). To add diffused residues to either side of this motif you can specify a range or discrete value without a chain letter infront. To add 15 residues before the motif and 20-30 residues (randomly sampled) after use the following syntax: 15,A3-10,20-30 commas are used to separate regions selected from the pdb and designed (diffused) resiudes which will be added. ')
|
| 482 |
-
pdb_id_code = gr.Textbox(label="PDB ID", lines=1, placeholder='INPUT PDB ID TO FETCH (ex. 1DPX)', visible=True)
|
| 483 |
-
contigs = gr.Textbox(label="contigs", lines=1, placeholder='specify contigs to grab particular residues from pdb ()', visible=True)
|
| 484 |
-
gr.Markdown('Using the same contig syntax, seq or str of input motif residues can be masked, allowing the model to hold strucutre fixed and design sequence or vice-versa')
|
| 485 |
-
with gr.Row():
|
| 486 |
-
seq_mask = gr.Textbox(label='seq mask',lines=1,placeholder='input residues to mask sequence')
|
| 487 |
-
str_mask = gr.Textbox(label='str mask',lines=1,placeholder='input residues to mask structure')
|
| 488 |
-
preview_viewer = gr.HTML()
|
| 489 |
-
rewrite_pdb = gr.File(label='PDB file')
|
| 490 |
-
preview_btn = gr.Button("Preview Motif")
|
| 491 |
-
|
| 492 |
-
with gr.TabItem("MSA to PSSM"):
|
| 493 |
-
gr.Markdown("""### MSA to PSSM Generation""")
|
| 494 |
-
gr.Markdown('input either an MSA or PSSM to guide the model toward generating samples within your family of interest')
|
| 495 |
-
with gr.Row():
|
| 496 |
-
fasta_msa = gr.File(label='MSA')
|
| 497 |
-
input_pssm = gr.File(label='PSSM (.csv)')
|
| 498 |
-
pssm = gr.File(label='Generated PSSM')
|
| 499 |
-
pssm_view = gr.Plot(label='PSSM Viewer')
|
| 500 |
-
pssm_gen_btn = gr.Button("Generate PSSM")
|
| 501 |
-
|
| 502 |
-
|
| 503 |
-
btn = gr.Button("GENERATE")
|
| 504 |
-
|
| 505 |
-
#with gr.Row():
|
| 506 |
-
with gr.Column():
|
| 507 |
-
gr.Markdown("""## OUTPUTS""")
|
| 508 |
-
gr.Markdown("""#### Confidence score for generated structure at each timestep""")
|
| 509 |
-
plddt_plot = gr.Plot(label='plddt at step t')
|
| 510 |
-
gr.Markdown("""#### Output protein sequnece""")
|
| 511 |
-
output_seq = gr.Textbox(label="sequence")
|
| 512 |
-
gr.Markdown("""#### Download PDB file""")
|
| 513 |
-
output_pdb = gr.File(label="PDB file")
|
| 514 |
-
gr.Markdown("""#### Structure viewer""")
|
| 515 |
-
output_viewer = gr.HTML()
|
| 516 |
-
'''
|
| 517 |
-
gr.Markdown("""### Don't know where to get started? Click on an example below to try it out!""")
|
| 518 |
-
gr.Examples(
|
| 519 |
-
[["","125",0.0,0.0,0.2,"","","","20","normal",'','','',None,'','',None],
|
| 520 |
-
["","100",0.0,0.0,0.0,"","W0.2","2","20","normal",'','','',None,'','',None],
|
| 521 |
-
# ["","100",0.0,0.0,0.0,
|
| 522 |
-
# "XXHHHHHHHHHXXXXXXXHHHHHHHHHXXXXXXXHHHHHHHHXXXXSSSSSSSSSSSXXXXXXXXSSSSSSSSSSSSXXXXXXXSSSSSSSSSXXXXXXX",
|
| 523 |
-
# "","","25","normal",'','','',None,'','',None],
|
| 524 |
-
# ["XXXXXXXXXXXXXXXXXXXXXXXXXIPDXXXXXXXXXXXXXXXXXXXXXXPEPSEQXXXXXXXXXXXXXXXXXXXXXXXXXXIPDXXXXXXXXXXXXXXXXXXX",
|
| 525 |
-
# "",0.0,0.0,0.0,"","","","25","normal",'','','',None,'','',None],
|
| 526 |
-
# ["","",0.0,0.0,0.0,"","","","25","normal",'','',
|
| 527 |
-
# '9,D10-11,8,D20-20,4,D25-35,65,D101-101,2,D104-105,8,D114-116,15,D132-138,6,D145-145,2,D148-148,12,D161-161,3',
|
| 528 |
-
# './tmp/PSSM_lysozyme.csv',
|
| 529 |
-
# 'D25-25,D27-31,D33-35,D132-137',
|
| 530 |
-
# 'D26-26','./tmp/150l.pdb']
|
| 531 |
-
],
|
| 532 |
-
inputs=[sequence,
|
| 533 |
-
seq_len,
|
| 534 |
-
helix_bias,
|
| 535 |
-
strand_bias,
|
| 536 |
-
loop_bias,
|
| 537 |
-
secondary_structure,
|
| 538 |
-
aa_bias,
|
| 539 |
-
aa_bias_potential,
|
| 540 |
-
#target_charge,
|
| 541 |
-
#target_ph,
|
| 542 |
-
#charge_potential,
|
| 543 |
-
num_steps,
|
| 544 |
-
noise,
|
| 545 |
-
hydrophobic_target_score,
|
| 546 |
-
hydrophobic_potential,
|
| 547 |
-
contigs,
|
| 548 |
-
pssm,
|
| 549 |
-
seq_mask,
|
| 550 |
-
str_mask,
|
| 551 |
-
rewrite_pdb],
|
| 552 |
-
outputs=[output_seq,
|
| 553 |
-
output_pdb,
|
| 554 |
-
output_viewer,
|
| 555 |
-
plddt_plot],
|
| 556 |
-
fn=protein_diffusion_model,
|
| 557 |
-
)
|
| 558 |
-
'''
|
| 559 |
-
preview_btn.click(get_motif_preview,[pdb_id_code, contigs],[preview_viewer, rewrite_pdb])
|
| 560 |
-
|
| 561 |
-
pssm_gen_btn.click(get_pssm,[fasta_msa,input_pssm],[pssm_view, pssm])
|
| 562 |
-
|
| 563 |
-
btn.click(protein_diffusion_model,
|
| 564 |
-
[sequence,
|
| 565 |
-
seq_len,
|
| 566 |
-
helix_bias,
|
| 567 |
-
strand_bias,
|
| 568 |
-
loop_bias,
|
| 569 |
-
secondary_structure,
|
| 570 |
-
aa_bias,
|
| 571 |
-
aa_bias_potential,
|
| 572 |
-
#target_charge,
|
| 573 |
-
#target_ph,
|
| 574 |
-
#charge_potential,
|
| 575 |
-
num_steps,
|
| 576 |
-
noise,
|
| 577 |
-
hydrophobic_target_score,
|
| 578 |
-
hydrophobic_potential,
|
| 579 |
-
contigs,
|
| 580 |
-
pssm,
|
| 581 |
-
seq_mask,
|
| 582 |
-
str_mask,
|
| 583 |
-
rewrite_pdb],
|
| 584 |
-
[output_seq,
|
| 585 |
-
output_pdb,
|
| 586 |
-
output_viewer,
|
| 587 |
-
plddt_plot])
|
| 588 |
-
|
| 589 |
-
demo.queue()
|
| 590 |
-
demo.launch(debug=True)
|
| 591 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|