Spaces:
Sleeping
Sleeping
Commit
•
e8b587f
1
Parent(s):
c2e36d2
fix button
Browse files
app.py
CHANGED
@@ -490,266 +490,98 @@ if selected_model == 'Cas9':
|
|
490 |
st.experimental_rerun()
|
491 |
|
492 |
elif selected_model == 'Cas12':
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
493 |
cas12target_selection = st.radio(
|
494 |
-
"Select either
|
495 |
('regular', 'mutation'),
|
496 |
key='cas12target_selection'
|
497 |
)
|
498 |
if 'current_gene_symbol' not in st.session_state:
|
499 |
st.session_state['current_gene_symbol'] = ""
|
500 |
|
501 |
-
# Define a function to clean up old files
|
502 |
-
|
503 |
def clean_up_old_files(gene_symbol):
|
504 |
-
|
505 |
-
|
506 |
-
|
507 |
-
|
508 |
-
|
509 |
-
|
510 |
-
|
511 |
-
|
512 |
-
|
513 |
-
|
|
|
514 |
|
515 |
-
|
516 |
-
if gene_symbol != st.session_state['current_gene_symbol'] and gene_symbol:
|
517 |
if st.session_state['current_gene_symbol']:
|
518 |
-
# Clean up files only if a different gene symbol is entered and a previous symbol exists
|
519 |
clean_up_old_files(st.session_state['current_gene_symbol'])
|
520 |
-
# Update the session state with the new gene symbol
|
521 |
st.session_state['current_gene_symbol'] = gene_symbol
|
522 |
|
523 |
-
|
524 |
-
|
525 |
-
|
526 |
-
|
527 |
-
|
528 |
-
|
529 |
-
|
530 |
-
|
531 |
-
|
532 |
-
|
533 |
-
|
534 |
-
|
535 |
-
|
536 |
-
st.session_state['exons'] = exons # Store exon data
|
537 |
-
|
538 |
-
# Notify the user once the process is completed successfully.
|
539 |
-
st.success('Prediction completed!')
|
540 |
-
st.session_state['prediction_made'] = True
|
541 |
-
|
542 |
-
if 'on_target_results' in st.session_state and st.session_state['on_target_results']:
|
543 |
-
ensembl_id = gene_annotations.get(gene_symbol, 'Unknown') # Get Ensembl ID or default to 'Unknown'
|
544 |
-
col1, col2, col3 = st.columns(3)
|
545 |
-
with col1:
|
546 |
-
st.markdown("**Genome**")
|
547 |
-
st.markdown("Homo sapiens")
|
548 |
-
with col2:
|
549 |
-
st.markdown("**Gene**")
|
550 |
-
st.markdown(f"{gene_symbol} : {ensembl_id} (primary)")
|
551 |
-
with col3:
|
552 |
-
st.markdown("**Nuclease**")
|
553 |
-
st.markdown("SpCas9")
|
554 |
-
# Include "Target" in the DataFrame's columns
|
555 |
-
try:
|
556 |
-
df = pd.DataFrame(st.session_state['on_target_results'],
|
557 |
-
columns=["Chr", "Start Pos", "End Pos", "Strand", "Transcript", "Exon",
|
558 |
-
"Target",
|
559 |
-
"gRNA", "Prediction"])
|
560 |
-
st.dataframe(df)
|
561 |
-
except ValueError as e:
|
562 |
-
st.error(f"DataFrame creation error: {e}")
|
563 |
-
# Optionally print or log the problematic data for debugging:
|
564 |
-
print(st.session_state['on_target_results'])
|
565 |
-
|
566 |
-
# Initialize Plotly figure
|
567 |
-
fig = go.Figure()
|
568 |
-
|
569 |
-
EXON_BASE = 0 # Base position for exons and CDS on the Y axis
|
570 |
-
EXON_HEIGHT = 0.02 # How 'tall' the exon markers should appear
|
571 |
-
|
572 |
-
# Plot Exons as small markers on the X-axis
|
573 |
-
for exon in st.session_state['exons']:
|
574 |
-
exon_start, exon_end = exon['start'], exon['end']
|
575 |
-
fig.add_trace(go.Bar(
|
576 |
-
x=[(exon_start + exon_end) / 2],
|
577 |
-
y=[EXON_HEIGHT],
|
578 |
-
width=[exon_end - exon_start],
|
579 |
-
base=EXON_BASE,
|
580 |
-
marker_color='rgba(128, 0, 128, 0.5)',
|
581 |
-
name='Exon'
|
582 |
-
))
|
583 |
-
|
584 |
-
VERTICAL_GAP = 0.2 # Gap between different ranks
|
585 |
-
|
586 |
-
# Define max and min Y values based on strand and rank
|
587 |
-
MAX_STRAND_Y = 0.1 # Maximum Y value for positive strand results
|
588 |
-
MIN_STRAND_Y = -0.1 # Minimum Y value for negative strand results
|
589 |
-
|
590 |
-
# Iterate over top 5 sorted predictions to create the plot
|
591 |
-
for i, prediction in enumerate(st.session_state['on_target_results'][:5], start=1): # Only top 5
|
592 |
-
chrom, start, end, strand, transcript, exon, target, gRNA, prediction_score = prediction
|
593 |
-
midpoint = (int(start) + int(end)) / 2
|
594 |
-
|
595 |
-
# Vertical position based on rank, modified by strand
|
596 |
-
y_value = (MAX_STRAND_Y - (i - 1) * VERTICAL_GAP) if strand == '1' or strand == '+' else (
|
597 |
-
MIN_STRAND_Y + (i - 1) * VERTICAL_GAP)
|
598 |
-
|
599 |
-
fig.add_trace(go.Scatter(
|
600 |
-
x=[midpoint],
|
601 |
-
y=[y_value],
|
602 |
-
mode='markers+text',
|
603 |
-
marker=dict(symbol='triangle-up' if strand == '1' or strand == '+' else 'triangle-down',
|
604 |
-
size=12),
|
605 |
-
text=f"Rank: {i}", # Text label
|
606 |
-
hoverinfo='text',
|
607 |
-
hovertext=f"Rank: {i}<br>Chromosome: {chrom}<br>Target Sequence: {target}<br>gRNA: {gRNA}<br>Start: {start}<br>End: {end}<br>Strand: {'+' if strand == '1' or strand == '+' else '-'}<br>Transcript: {transcript}<br>Prediction: {prediction_score:.4f}",
|
608 |
-
))
|
609 |
-
|
610 |
-
# Update layout for clarity and interaction
|
611 |
-
fig.update_layout(
|
612 |
-
title='Top 5 gRNA Sequences by Prediction Score',
|
613 |
-
xaxis_title='Genomic Position',
|
614 |
-
yaxis_title='Strand',
|
615 |
-
yaxis=dict(tickvals=[MAX_STRAND_Y, MIN_STRAND_Y], ticktext=['+', '-']),
|
616 |
-
showlegend=False,
|
617 |
-
hovermode='x unified',
|
618 |
-
)
|
619 |
-
|
620 |
-
# Display the plot
|
621 |
-
st.plotly_chart(fig)
|
622 |
-
|
623 |
-
if 'gene_sequence' in st.session_state and st.session_state['gene_sequence']:
|
624 |
-
gene_symbol = st.session_state['current_gene_symbol']
|
625 |
-
gene_sequence = st.session_state['gene_sequence']
|
626 |
-
|
627 |
-
# Define file paths
|
628 |
-
genbank_file_path = f"{gene_symbol}_crispr_targets.gb"
|
629 |
-
bed_file_path = f"{gene_symbol}_crispr_targets.bed"
|
630 |
-
csv_file_path = f"{gene_symbol}_crispr_predictions.csv"
|
631 |
-
plot_image_path = f"{gene_symbol}_gtracks_plot.png"
|
632 |
-
|
633 |
-
# Generate files
|
634 |
-
cas12lstm.generate_genbank_file_from_df(df, gene_sequence, gene_symbol, genbank_file_path)
|
635 |
-
cas12lstm.create_bed_file_from_df(df, bed_file_path)
|
636 |
-
cas12lstm.create_csv_from_df(df, csv_file_path)
|
637 |
-
|
638 |
-
# Prepare an in-memory buffer for the ZIP file
|
639 |
-
zip_buffer = io.BytesIO()
|
640 |
-
with zipfile.ZipFile(zip_buffer, 'w', zipfile.ZIP_DEFLATED) as zip_file:
|
641 |
-
# For each file, add it to the ZIP file
|
642 |
-
zip_file.write(genbank_file_path)
|
643 |
-
zip_file.write(bed_file_path)
|
644 |
-
zip_file.write(csv_file_path)
|
645 |
-
|
646 |
-
# Important: move the cursor to the beginning of the BytesIO buffer before reading it
|
647 |
-
zip_buffer.seek(0)
|
648 |
-
|
649 |
-
# Specify the region you want to visualize
|
650 |
-
min_start = df['Start Pos'].min()
|
651 |
-
max_end = df['End Pos'].max()
|
652 |
-
chromosome = df['Chr'].mode()[0] # Assumes most common chromosome is the target
|
653 |
-
region = f"{chromosome}:{min_start}-{max_end}"
|
654 |
-
|
655 |
-
# Generate the pyGenomeTracks plot
|
656 |
-
gtracks_command = f"gtracks {region} {bed_file_path} {plot_image_path}"
|
657 |
-
subprocess.run(gtracks_command, shell=True)
|
658 |
-
st.image(plot_image_path)
|
659 |
-
|
660 |
-
# Display the download button for the ZIP file
|
661 |
-
st.download_button(
|
662 |
-
label="Download GenBank, BED, CSV files as ZIP",
|
663 |
-
data=zip_buffer.getvalue(),
|
664 |
-
file_name=f"{gene_symbol}_files.zip",
|
665 |
-
mime="application/zip"
|
666 |
-
)
|
667 |
-
elif cas12target_selection == 'mutation':
|
668 |
-
# Prediction button
|
669 |
-
predict_button = st.button('Predict cas12')
|
670 |
-
vcf_reader = cyvcf2.VCF('SRR25934512.filter.snps.indels.vcf.gz')
|
671 |
-
|
672 |
-
if 'exons' not in st.session_state:
|
673 |
-
st.session_state['exons'] = []
|
674 |
-
|
675 |
-
# Process predictions
|
676 |
-
if predict_button and gene_symbol:
|
677 |
-
with st.spinner('Predicting... Please wait'):
|
678 |
-
predictions, gene_sequence, exons = cas12lstmvcf.process_gene(gene_symbol, vcf_reader,
|
679 |
-
cas12lstm_path)
|
680 |
-
full_predictions = sorted(predictions, key=lambda x: x[8], reverse=True)
|
681 |
-
sorted_predictions = sorted(predictions, key=lambda x: x[8], reverse=True)[:10]
|
682 |
-
st.session_state['full_results'] = full_predictions
|
683 |
-
st.session_state['on_target_results'] = sorted_predictions
|
684 |
-
st.session_state['gene_sequence'] = gene_sequence # Save gene sequence in session state
|
685 |
-
st.session_state['exons'] = exons # Store exon data
|
686 |
-
|
687 |
-
# Notify the user once the process is completed successfully.
|
688 |
-
st.success('Prediction completed!')
|
689 |
-
st.session_state['prediction_made'] = True
|
690 |
-
|
691 |
-
if 'on_target_results' in st.session_state and st.session_state['on_target_results']:
|
692 |
-
ensembl_id = gene_annotations.get(gene_symbol,
|
693 |
-
'Unknown') # Get Ensembl ID or default to 'Unknown'
|
694 |
-
col1, col2, col3 = st.columns(3)
|
695 |
-
with col1:
|
696 |
-
st.markdown("**Genome**")
|
697 |
-
st.markdown("Homo sapiens")
|
698 |
-
with col2:
|
699 |
-
st.markdown("**Gene**")
|
700 |
-
st.markdown(f"{gene_symbol} : {ensembl_id} (primary)")
|
701 |
-
with col3:
|
702 |
-
st.markdown("**Nuclease**")
|
703 |
-
st.markdown("SpCas9")
|
704 |
-
# Include "Target" in the DataFrame's columns
|
705 |
-
try:
|
706 |
-
df = pd.DataFrame(st.session_state['on_target_results'],
|
707 |
-
columns=["Gene Symbol", "Chr", "Strand", "Target Start", "Transcript",
|
708 |
-
"Exon",
|
709 |
-
"Target",
|
710 |
-
"gRNA", "Prediction", "Is Mutation"])
|
711 |
-
df_full = pd.DataFrame(st.session_state['full_results'],
|
712 |
-
columns=["Gene Symbol", "Chr", "Strand", "Target Start",
|
713 |
-
"Transcript",
|
714 |
-
"Exon", "Target",
|
715 |
-
"gRNA", "Prediction", "Is Mutation"])
|
716 |
-
st.dataframe(df)
|
717 |
-
except ValueError as e:
|
718 |
-
st.error(f"DataFrame creation error: {e}")
|
719 |
-
# Optionally print or log the problematic data for debugging:
|
720 |
-
print(st.session_state['on_target_results'])
|
721 |
-
|
722 |
-
if 'gene_sequence' in st.session_state and st.session_state['gene_sequence']:
|
723 |
-
gene_symbol = st.session_state['current_gene_symbol']
|
724 |
-
gene_sequence = st.session_state['gene_sequence']
|
725 |
-
|
726 |
-
# Define file paths
|
727 |
-
genbank_file_path = f"{gene_symbol}_crispr_targets.gb"
|
728 |
-
bed_file_path = f"{gene_symbol}_crispr_targets.bed"
|
729 |
-
csv_file_path = f"{gene_symbol}_crispr_predictions.csv"
|
730 |
-
plot_image_path = f"{gene_symbol}_gtracks_plot.png"
|
731 |
-
|
732 |
-
# Generate files
|
733 |
-
cas12lstmvcf.generate_genbank_file_from_df(df_full, gene_sequence, gene_symbol,
|
734 |
-
genbank_file_path)
|
735 |
-
cas12lstmvcf.create_bed_file_from_df(df_full, bed_file_path)
|
736 |
-
cas12lstmvcf.create_csv_from_df(df_full, csv_file_path)
|
737 |
-
|
738 |
-
# Prepare an in-memory buffer for the ZIP file
|
739 |
-
zip_buffer = io.BytesIO()
|
740 |
-
with zipfile.ZipFile(zip_buffer, 'w', zipfile.ZIP_DEFLATED) as zip_file:
|
741 |
-
# For each file, add it to the ZIP file
|
742 |
-
zip_file.write(genbank_file_path)
|
743 |
-
zip_file.write(bed_file_path)
|
744 |
-
zip_file.write(csv_file_path)
|
745 |
-
|
746 |
-
# Display the download button for the ZIP file
|
747 |
-
st.download_button(
|
748 |
-
label="Download GenBank, BED, CSV files as ZIP",
|
749 |
-
data=zip_buffer.getvalue(),
|
750 |
-
file_name=f"{gene_symbol}_files.zip",
|
751 |
-
mime="application/zip"
|
752 |
-
)
|
753 |
|
754 |
elif selected_model == 'Cas13d':
|
755 |
ENTRY_METHODS = dict(
|
|
|
490 |
st.experimental_rerun()
|
491 |
|
492 |
elif selected_model == 'Cas12':
|
493 |
+
def visualize_and_generate_files(df, gene_sequence, exons, gene_symbol):
|
494 |
+
fig = go.Figure()
|
495 |
+
# Exon visualization
|
496 |
+
for exon in exons:
|
497 |
+
exon_start, exon_end = exon['start'], exon['end']
|
498 |
+
fig.add_trace(go.Bar(x=[(exon_start + exon_end) / 2], y=[0.5], width=[exon_end - exon_start], base=0,
|
499 |
+
marker_color='purple', name='Exon'))
|
500 |
+
# Prediction visualization
|
501 |
+
for i, prediction in enumerate(df.itertuples(), start=1):
|
502 |
+
fig.add_trace(go.Scatter(x=[(prediction.Start_Pos + prediction.End_Pos) / 2], y=[1], mode='markers',
|
503 |
+
marker=dict(size=10, color='blue'), name=f'Prediction {i}'))
|
504 |
+
fig.update_layout(title='Cas12 Prediction Visualization', xaxis_title='Position',
|
505 |
+
yaxis=dict(tickvals=[0.5, 1], ticktext=['Exons', 'Predictions']), showlegend=True)
|
506 |
+
st.plotly_chart(fig)
|
507 |
+
|
508 |
+
# File generation and download
|
509 |
+
generate_and_download_files(df, gene_symbol)
|
510 |
+
|
511 |
+
|
512 |
+
def generate_and_download_files(df, gene_symbol):
|
513 |
+
genbank_file_path = f"{gene_symbol}_crispr_targets.gb"
|
514 |
+
bed_file_path = f"{gene_symbol}_crispr_targets.bed"
|
515 |
+
csv_file_path = f"{gene_symbol}_crispr_predictions.csv"
|
516 |
+
df.to_csv(csv_file_path, index=False)
|
517 |
+
# Assume functions to generate GenBank and BED are defined in cas12lstm or cas12lstmvcf
|
518 |
+
cas12lstm.generate_genbank_file_from_df(df, gene_symbol, genbank_file_path)
|
519 |
+
cas12lstm.create_bed_file_from_df(df, bed_file_path)
|
520 |
+
|
521 |
+
zip_buffer = io.BytesIO()
|
522 |
+
with zipfile.ZipFile(zip_buffer, 'w', zipfile.ZIP_DEFLATED) as zip_file:
|
523 |
+
zip_file.write(genbank_file_path)
|
524 |
+
zip_file.write(bed_file_path)
|
525 |
+
zip_file.write(csv_file_path)
|
526 |
+
zip_buffer.seek(0)
|
527 |
+
st.download_button("Download GenBank, BED, CSV files as ZIP", data=zip_buffer.getvalue(),
|
528 |
+
file_name=f"{gene_symbol}_files.zip", mime="application/zip")
|
529 |
+
|
530 |
+
|
531 |
+
def display_results(predictions, gene_sequence, exons, gene_symbol):
|
532 |
+
st.success('Prediction completed!')
|
533 |
+
ensembl_id = gene_annotations.get(gene_symbol, 'Unknown')
|
534 |
+
st.write(f"**Genome:** Homo sapiens")
|
535 |
+
st.write(f"**Gene:** {gene_symbol} : {ensembl_id} (primary)")
|
536 |
+
st.write("**Nuclease:** Cas12")
|
537 |
+
df = pd.DataFrame(predictions,
|
538 |
+
columns=["Chr", "Start Pos", "End Pos", "Strand", "Transcript", "Exon", "Target", "gRNA",
|
539 |
+
"Prediction"])
|
540 |
+
st.dataframe(df)
|
541 |
+
|
542 |
+
# Visualization and file generation as demonstrated in the Cas9 example
|
543 |
+
visualize_and_generate_files(df, gene_sequence, exons, gene_symbol)
|
544 |
+
|
545 |
+
|
546 |
cas12target_selection = st.radio(
|
547 |
+
"Select either regular or mutation:",
|
548 |
('regular', 'mutation'),
|
549 |
key='cas12target_selection'
|
550 |
)
|
551 |
if 'current_gene_symbol' not in st.session_state:
|
552 |
st.session_state['current_gene_symbol'] = ""
|
553 |
|
|
|
|
|
554 |
def clean_up_old_files(gene_symbol):
|
555 |
+
for suffix in ['_crispr_targets.gb', '_crispr_targets.bed', '_crispr_predictions.csv']:
|
556 |
+
file_path = f"{gene_symbol}{suffix}"
|
557 |
+
if os.path.exists(file_path):
|
558 |
+
os.remove(file_path)
|
559 |
+
|
560 |
+
gene_symbol = st.selectbox(
|
561 |
+
'Enter a Gene Symbol:',
|
562 |
+
[''] + gene_symbol_list,
|
563 |
+
key='gene_symbol',
|
564 |
+
format_func=lambda x: x if x else ""
|
565 |
+
)
|
566 |
|
567 |
+
if gene_symbol != st.session_state['current_gene_symbol']:
|
|
|
568 |
if st.session_state['current_gene_symbol']:
|
|
|
569 |
clean_up_old_files(st.session_state['current_gene_symbol'])
|
|
|
570 |
st.session_state['current_gene_symbol'] = gene_symbol
|
571 |
|
572 |
+
if cas12target_selection == 'regular':
|
573 |
+
if st.button('Predict cas12 Regular'):
|
574 |
+
with st.spinner('Predicting... Please wait'):
|
575 |
+
predictions, gene_sequence, exons = cas12lstm.process_gene(gene_symbol, cas12lstm_path)
|
576 |
+
sorted_predictions = sorted(predictions, key=lambda x: x[8], reverse=True)[:10]
|
577 |
+
display_results(sorted_predictions, gene_sequence, exons, gene_symbol)
|
578 |
+
elif cas12target_selection == 'mutation':
|
579 |
+
vcf_reader = cyvcf2.VCF('SRR25934512.filter.snps.indels.vcf.gz')
|
580 |
+
if st.button('Predict cas12 Mutation'):
|
581 |
+
with st.spinner('Predicting... Please wait'):
|
582 |
+
predictions, gene_sequence, exons = cas12lstmvcf.process_gene(gene_symbol, vcf_reader, cas12lstm_path)
|
583 |
+
sorted_predictions = sorted(predictions, key=lambda x: x[8], reverse=True)[:10]
|
584 |
+
display_results(sorted_predictions, gene_sequence, exons, gene_symbol)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
585 |
|
586 |
elif selected_model == 'Cas13d':
|
587 |
ENTRY_METHODS = dict(
|