Upload folder using huggingface_hub

README.md

@@ -0,0 +1,120 @@
+---
+license: mit
+library_name: chrombpnet
+tags:
+  - encode
+  - chrombpnet
+  - chromatin-accessibility
+  - ATAC
+  - lung
+  - hg38
+---
+# ENCODE ChromBPNet Atlas
+As part of the ENCODE 4 Project, we trained ChromBPNet models on 1,512 ENCODE DNAse-seq and ATAC-seq across 408 biosamples. Here, we provide all models for open-source use.
+
+For more information about the models, see:
+- Main ENCODE 4 Paper
+- [A unified lexicon of predictive DNA sequence motifs from ENCODE transcription factor binding and chromatin accessibility assays](https://doi.org/10.5281/zenodo.17123347) (Deshpande et al., Zenodo 2025)
+- [ChromBPNet: bias factorized, base-resolution deep learning models of chromatin accessibility reveal cis-regulatory sequence syntax, transcription factor footprints and regulatory variants](https://doi.org/10.1101/2024.12.25.630221) (Pampari et al., bioRxiv 2024)
+
+## ChromBPNet model: ATAC in lower lobe of right lung (ENCSR923VTG)
+- Model: ChromBPNet
+- Assay: ATAC-seq
+- Experiment: [ENCSR923VTG](https://www.encodeproject.org/experiments/ENCSR923VTG/)
+- Model annotation: [ENCSR609WOB](https://www.encodeproject.org/annotations/ENCSR609WOB/)
+- Biosample: lower lobe of right lung (Full name: Homo sapiens lower lobe of right lung tissue male adult (60 years))
+- Cell slim(s): None
+- Organ slim(s): lung
+- Developmental slim(s): endoderm
+- System slim(s): respiratory-system
+- Assembly: hg38
+
+## Directory structure
+- `fold_0`: Model of 5-fold cross-validation: Fold 0
+    - `model.chrombpnet.fold_0.encid.h5`: full chrombpnet model that combines both bias and corrected model in .h5 format
+    - `model.chrombpnet_nobias.fold_0.encid.h5`: bias-corrected accessibility model in .h5 format (Use for all biological discovery)
+    - `model.bias_scaled.fold_0.encid.h5`: bias model in .h5 format
+    - `model.chrombpnet.fold_0.encid.tar`: full chrombpnet model that combines both bias and  corrected model in SavedModel format. After being untarred, it results in a directory named "chrombpnet".
+    - `model.chrombpnet_nobias.fold_0.encid.tar`: bias-corrected accessibility model in SavedModel format (Use for all biological discovery). After being untarred, it results in a directory named "chrombpnet_wo_bias".
+    - `model.bias_scaled.fold_0.encid.tar`: bias model in SavedModel format. After being untarred, it results in a directory named "bias_model_scaled".
+    - `logs.models.fold_0.encid`: folder containing log files for training models
+- `fold_1`: Model of 5-fold coss-validation: Fold 1
+- `fold_2`: Model of 5-fold cross-validation: Fold 2
+- `fold_3`: Model of 5-fold cross-validation: Fold 3
+- `fold_4`: Model of 5-fold cross-validation: Fold 4
+
+# Instructions
+## 1. Pseudocode for loading models in .h5 format 
+
+(1) Use the code in python after appropriately defining `model_in_h5_format` and `inputs`. \
+(2) `inputs` is a one hot encoded sequence of shape (N,2114,4). Here N corresponds to the 
+number of tested sequences, 2114 is the input sequence length and 4 corresponds to [A,C,G,T].
+
+```python
+import tensorflow as tf
+from tensorflow.keras.utils import get_custom_objects
+from tensorflow.keras.models import load_model
+
+custom_objects={"tf": tf}
+get_custom_objects().update(custom_objects)
+
+model=load_model(model_in_h5_format,compile=False)
+outputs = model(inputs)
+```
+
+The list `outputs` consists of two elements. The first element has a shape of (N, 1000) and
+contains logit predictions for a 1000-base-pair output. The second element, with a shape of
+(N, 1), contains logcount predictions. To transform these predictions into per-base signals, 
+follow the provided pseudo code lines below.
+
+```python
+import numpy as np
+
+def softmax(x, temp=1):
+    norm_x = x - np.mean(x,axis=1, keepdims=True)
+    return np.exp(temp*norm_x)/np.sum(np.exp(temp*norm_x), axis=1, keepdims=True)
+    
+predictions = softmax(outputs[0]) * (np.exp(outputs[1])-1)
+```
+
+## 2. Pseudocode for loading models in .tar format
+
+(1) First untar the directory as follows `tar -xvf model.tar`. \
+(2) Use the code below in python after appropriately defining `model_dir_untared` and `inputs`. \
+(3) `inputs` is a one hot encoded sequence of shape (N,2114,4). Here N corresponds to the number
+of tested sequences, 2114 is the input sequence length and 4 corresponds to ACGT.
+
+Reference: https://www.tensorflow.org/api_docs/python/tf/saved_model/load
+
+```python
+import tensorflow as tf
+
+model = tf.saved_model.load('model_dir_untared')
+outputs = model.signatures['serving_default'](**{'sequence':inputs.astype('float32')})
+```
+
+The variable `outputs` represents a dictionary containing two key-value pairs. The first key
+is `logits_profile_predictions`, holding a value with a shape of (N, 1000). This value corresponds
+to logit predictions for a 1000-base-pair output. The second key, named `logcount_predictions``, 
+is associated with a value of shape (N, 1), representing logcount predictions. To transform these
+predictions into per-base signals, utilize the provided pseudo code lines mentioned below.
+
+```python
+import numpy as np
+def softmax(x, temp=1):
+    norm_x = x - np.mean(x,axis=1, keepdims=True)
+    return np.exp(temp*norm_x)/np.sum(np.exp(temp*norm_x), axis=1, keepdims=True)
+    
+predictions = softmax(outputs["logits_profile_predictions"]) * (np.exp(outputs["logcount_predictions"])-1)
+```
+
+## Docker image to load and use the models
+- https://hub.docker.com/r/kundajelab/chrombpnet-atlas/ (tag:v1)
+
+## Code for ChromBPNet
+- https://github.com/kundajelab/chrombpnet/
+
+# License & citation
+External data users may freely download, analyze and publish results based on any ENCODE data without restrictions.
+
+Released under the [ENCODE data-use policy](https://www.encodeproject.org/about/data-use-policy/). Please cite the ENCODE Project Consortium and the model software: [ChromBPNet](https://github.com/kundajelab/chrombpnet) (Pampari et al., bioRxiv 2024).

fold_0/logs.models.fold_0.ENCSR923VTG/logfile.modelling.fold_0.ENCSR923VTG.args.json

@@ -0,0 +1,50 @@
+{
+    "cmd": "pipeline",
+    "genome": "/oak/stanford/groups/akundaje/ziwei75/atac_seq_pipeline/hg38/GRCh38_no_alt_analysis_set_GCA_000001405.15.fasta",
+    "chrom_sizes": "/oak/stanford/groups/akundaje/ziwei75/atac_seq_pipeline/hg38/GRCh38_EBV.chrom.sizes.tsv",
+    "input_bam_file": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/preprocessing/bigWigs/ENCSR923VTG.bigWig",
+    "input_fragment_file": null,
+    "input_tagalign_file": null,
+    "output_dir": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_0/model_outputs",
+    "data_type": "ATAC",
+    "peaks": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_0/model_outputs/auxiliary/filtered.peaks.bed",
+    "nonpeaks": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_0/model_outputs/auxiliary/filtered.nonpeaks.bed",
+    "chr_fold_path": "/oak/stanford/groups/akundaje/projects/chromatin-atlas-2022/splits/fold_0.json",
+    "outlier_threshold": 0.9999,
+    "ATAC_ref_path": null,
+    "DNASE_ref_path": null,
+    "num_samples": 10000,
+    "inputlen": 2114,
+    "outputlen": 1000,
+    "seed": 1234,
+    "epochs": 50,
+    "early_stop": 5,
+    "learning_rate": 0.001,
+    "trackables": [
+        "logcount_predictions_loss",
+        "loss",
+        "logits_profile_predictions_loss",
+        "val_logcount_predictions_loss",
+        "val_loss",
+        "val_logits_profile_predictions_loss"
+    ],
+    "architecture_from_file": "/home/users/vhecht/chrombpnet/chrombpnet/chrombpnet/training/models/chrombpnet_with_bias_model.py",
+    "file_prefix": null,
+    "html_prefix": "./",
+    "bsort": false,
+    "tmpdir": null,
+    "no_st": false,
+    "bias_model_path": "/scratch/users/vhecht/chromatin-atlas-2022/reference/ATAC/fold_0/model.bias.fold_0.ENCSR868FGK.h5",
+    "negative_sampling_ratio": 0.1,
+    "filters": 512,
+    "n_dilation_layers": 8,
+    "max_jitter": 500,
+    "batch_size": 64,
+    "output_prefix": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_0/model_outputs/models/chrombpnet",
+    "bigwig": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_0/model_outputs/auxiliary/data_unstranded.bw",
+    "plus_shift": null,
+    "minus_shift": null,
+    "chr": "chr8",
+    "pwm_width": 24,
+    "params": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_0/model_outputs/logs/chrombpnet_model_params.tsv"
+}

fold_0/logs.models.fold_0.ENCSR923VTG/logfile.modelling.fold_0.ENCSR923VTG.bias_formatting.stdout.txt

@@ -0,0 +1 @@
+Converting /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_0/model_outputs/models/bias_model_scaled.h5 to /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_0/new_model_format/bias_model_scaled.tar with get_new_tf_model_format.py

fold_0/logs.models.fold_0.ENCSR923VTG/logfile.modelling.fold_0.ENCSR923VTG.chrombpnet_data_params.tsv

@@ -0,0 +1,3 @@
+counts_sum_min_thresh	0.0
+counts_sum_max_thresh	10779.82
+trainings_pts_post_thresh	190682

fold_0/logs.models.fold_0.ENCSR923VTG/logfile.modelling.fold_0.ENCSR923VTG.chrombpnet_formatting.stdout.txt

@@ -0,0 +1 @@
+Converting /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_0/model_outputs/models/chrombpnet.h5 to /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_0/new_model_format/chrombpnet.tar with get_new_tf_model_format.py

fold_0/logs.models.fold_0.ENCSR923VTG/logfile.modelling.fold_0.ENCSR923VTG.chrombpnet_model_params.tsv

@@ -0,0 +1,9 @@
+counts_loss_weight	41.8
+filters	512
+n_dil_layers	8
+bias_model_path	/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_0/model_outputs/models/bias_model_scaled.h5
+inputlen	2114
+outputlen	1000
+max_jitter	500
+chr_fold_path	/oak/stanford/groups/akundaje/projects/chromatin-atlas-2022/splits/fold_0.json
+negative_sampling_ratio	0.1

fold_0/logs.models.fold_0.ENCSR923VTG/logfile.modelling.fold_0.ENCSR923VTG.chrombpnet_no_bias_formatting.stdout.txt

@@ -0,0 +1 @@
+Converting /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_0/model_outputs/models/chrombpnet_nobias.h5 to /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_0/new_model_format/chrombpnet_nobias.tar with get_new_tf_model_format.py

fold_0/logs.models.fold_0.ENCSR923VTG/logfile.modelling.fold_0.ENCSR923VTG.epoch_loss.csv

@@ -0,0 +1,14 @@
+epoch,logcount_predictions_loss,logits_profile_predictions_loss,loss,val_logcount_predictions_loss,val_logits_profile_predictions_loss,val_loss
+0,1.2342227697372437,684.477294921875,736.0678100585938,0.417678564786911,659.2682495117188,676.7271118164062
+1,0.3733308017253876,652.90966796875,668.5147705078125,0.3109683692455292,647.6123657226562,660.610595703125
+2,0.3419879674911499,642.7557373046875,657.05126953125,0.29698148369789124,642.9713745117188,655.3853149414062
+3,0.31577685475349426,635.1470336914062,648.345703125,0.27300769090652466,641.1583862304688,652.570068359375
+4,0.3001585602760315,628.764404296875,641.3103637695312,0.26966163516044617,639.1056518554688,650.3772583007812
+5,0.2828734517097473,623.91015625,635.734130859375,0.2639487087726593,639.0160522460938,650.0489501953125
+6,0.2710760235786438,619.5868530273438,630.918212890625,0.2550637125968933,638.237060546875,648.8994750976562
+7,0.2620992064476013,617.317138671875,628.272705078125,0.2547524869441986,638.1740112304688,648.8225708007812
+8,0.25144481658935547,613.6004638671875,624.1109619140625,0.29287001490592957,639.7479248046875,651.98974609375
+9,0.2444744110107422,612.1473999023438,622.3671264648438,0.24799196422100067,638.9130249023438,649.2794799804688
+10,0.23308557271957397,610.8416137695312,620.5834350585938,0.25178268551826477,640.017822265625,650.542236328125
+11,0.2245977520942688,609.2547607421875,618.6431274414062,0.2800882160663605,639.8231811523438,651.5308837890625
+12,0.21943892538547516,607.209228515625,616.380859375,0.24892912805080414,639.6738891601562,650.07958984375

fold_0/model.bias_scaled.fold_0.ENCSR923VTG.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b36a27ab4271dffa162a107dc4b037122ac8679cadd1da318ee73cbd43d84e5c
+size 2691928

fold_0/model.bias_scaled.fold_0.ENCSR923VTG.tar

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:58c356b2b77c40132b51f6bd13e12aa3625fabf1fbbe5ab258f6409a3d0d1cdd
+size 1198080

fold_0/model.chrombpnet.fold_0.ENCSR923VTG.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:725aa8c00b93963a2b92d2ed0765968dcbf5eb4260300de98ff1d0917ffe8fdc
+size 77538896

fold_0/model.chrombpnet.fold_0.ENCSR923VTG.tar

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:96f48102edf50be22c76188036f389dcd5b0083a70a8cdf8706b88da5cb8941c
+size 27535360

fold_0/model.chrombpnet_nobias.fold_0.ENCSR923VTG.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a4f92c67db2fa5425ffc6399c292bd5cbd243b9ab98c095ee9919e0ebe51568
+size 25582648

fold_0/model.chrombpnet_nobias.fold_0.ENCSR923VTG.tar

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d17360c2acb0bb4beac43762ae72f9fe45cd81cbf297e15d75d3da0ddd0a2248
+size 26060800

fold_1/logs.models.fold_1.ENCSR923VTG/logfile.modelling.fold_1.ENCSR923VTG.args.json

@@ -0,0 +1,50 @@
+{
+    "cmd": "pipeline",
+    "genome": "/oak/stanford/groups/akundaje/ziwei75/atac_seq_pipeline/hg38/GRCh38_no_alt_analysis_set_GCA_000001405.15.fasta",
+    "chrom_sizes": "/oak/stanford/groups/akundaje/ziwei75/atac_seq_pipeline/hg38/GRCh38_EBV.chrom.sizes.tsv",
+    "input_bam_file": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/preprocessing/bigWigs/ENCSR923VTG.bigWig",
+    "input_fragment_file": null,
+    "input_tagalign_file": null,
+    "output_dir": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_1/model_outputs",
+    "data_type": "ATAC",
+    "peaks": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_1/model_outputs/auxiliary/filtered.peaks.bed",
+    "nonpeaks": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_1/model_outputs/auxiliary/filtered.nonpeaks.bed",
+    "chr_fold_path": "/oak/stanford/groups/akundaje/projects/chromatin-atlas-2022/splits/fold_1.json",
+    "outlier_threshold": 0.9999,
+    "ATAC_ref_path": null,
+    "DNASE_ref_path": null,
+    "num_samples": 10000,
+    "inputlen": 2114,
+    "outputlen": 1000,
+    "seed": 1234,
+    "epochs": 50,
+    "early_stop": 5,
+    "learning_rate": 0.001,
+    "trackables": [
+        "logcount_predictions_loss",
+        "loss",
+        "logits_profile_predictions_loss",
+        "val_logcount_predictions_loss",
+        "val_loss",
+        "val_logits_profile_predictions_loss"
+    ],
+    "architecture_from_file": "/home/users/vhecht/chrombpnet/chrombpnet/chrombpnet/training/models/chrombpnet_with_bias_model.py",
+    "file_prefix": null,
+    "html_prefix": "./",
+    "bsort": false,
+    "tmpdir": null,
+    "no_st": false,
+    "bias_model_path": "/scratch/users/vhecht/chromatin-atlas-2022/reference/ATAC/fold_0/model.bias.fold_0.ENCSR868FGK.h5",
+    "negative_sampling_ratio": 0.1,
+    "filters": 512,
+    "n_dilation_layers": 8,
+    "max_jitter": 500,
+    "batch_size": 64,
+    "output_prefix": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_1/model_outputs/models/chrombpnet",
+    "bigwig": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_1/model_outputs/auxiliary/data_unstranded.bw",
+    "plus_shift": null,
+    "minus_shift": null,
+    "chr": "chr12",
+    "pwm_width": 24,
+    "params": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_1/model_outputs/logs/chrombpnet_model_params.tsv"
+}

fold_1/logs.models.fold_1.ENCSR923VTG/logfile.modelling.fold_1.ENCSR923VTG.bias_formatting.stdout.txt

@@ -0,0 +1 @@
+Converting /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_1/model_outputs/models/bias_model_scaled.h5 to /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_1/new_model_format/bias_model_scaled.tar with get_new_tf_model_format.py

fold_1/logs.models.fold_1.ENCSR923VTG/logfile.modelling.fold_1.ENCSR923VTG.chrombpnet_data_params.tsv

@@ -0,0 +1,3 @@
+counts_sum_min_thresh	0.0
+counts_sum_max_thresh	10651.59
+trainings_pts_post_thresh	192867

fold_1/logs.models.fold_1.ENCSR923VTG/logfile.modelling.fold_1.ENCSR923VTG.chrombpnet_formatting.stdout.txt

@@ -0,0 +1 @@
+Converting /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_1/model_outputs/models/chrombpnet.h5 to /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_1/new_model_format/chrombpnet.tar with get_new_tf_model_format.py

fold_1/logs.models.fold_1.ENCSR923VTG/logfile.modelling.fold_1.ENCSR923VTG.chrombpnet_model_params.tsv

@@ -0,0 +1,9 @@
+counts_loss_weight	41.8
+filters	512
+n_dil_layers	8
+bias_model_path	/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_1/model_outputs/models/bias_model_scaled.h5
+inputlen	2114
+outputlen	1000
+max_jitter	500
+chr_fold_path	/oak/stanford/groups/akundaje/projects/chromatin-atlas-2022/splits/fold_1.json
+negative_sampling_ratio	0.1

fold_1/logs.models.fold_1.ENCSR923VTG/logfile.modelling.fold_1.ENCSR923VTG.chrombpnet_no_bias_formatting.stdout.txt

@@ -0,0 +1 @@
+Converting /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_1/model_outputs/models/chrombpnet_nobias.h5 to /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_1/new_model_format/chrombpnet_nobias.tar with get_new_tf_model_format.py

fold_1/logs.models.fold_1.ENCSR923VTG/logfile.modelling.fold_1.ENCSR923VTG.epoch_loss.csv

@@ -0,0 +1,20 @@
+epoch,logcount_predictions_loss,logits_profile_predictions_loss,loss,val_logcount_predictions_loss,val_logits_profile_predictions_loss,val_loss
+0,1.3833131790161133,680.1495361328125,737.97216796875,0.5047076940536499,740.712158203125,761.8084716796875
+1,0.3850129246711731,648.7491455078125,664.8423461914062,0.3975262939929962,725.6054077148438,742.22216796875
+2,0.3474240303039551,637.967529296875,652.4905395507812,0.3348524570465088,718.0991821289062,732.0956420898438
+3,0.3216954469680786,629.7188110351562,643.1654052734375,0.3100779950618744,716.0814819335938,729.0432739257812
+4,0.2947860062122345,623.3861083984375,635.7081909179688,0.33821380138397217,715.8124389648438,729.9501342773438
+5,0.28720730543136597,617.8746337890625,629.87939453125,0.2811397314071655,710.46875,722.2203369140625
+6,0.27802610397338867,614.1626586914062,625.7832641601562,0.36926543712615967,713.6439819335938,729.079345703125
+7,0.2648470997810364,610.4195556640625,621.4906616210938,0.3466423451900482,711.2510986328125,725.7412109375
+8,0.24973343312740326,607.379638671875,617.8191528320312,0.27063924074172974,711.4103393554688,722.723388671875
+9,0.24141955375671387,605.0881958007812,615.1806030273438,0.2747054994106293,710.3330688476562,721.8154907226562
+10,0.2312697023153305,603.043212890625,612.7090454101562,0.27034828066825867,711.8262939453125,723.1270751953125
+11,0.22897033393383026,601.2933959960938,610.8651123046875,0.27081137895584106,711.2727661132812,722.5928955078125
+12,0.21682077646255493,600.0858764648438,609.1500244140625,0.28255191445350647,711.8565673828125,723.6674194335938
+13,0.2118193507194519,598.6821899414062,607.5354614257812,0.26435601711273193,710.4210815429688,721.4712524414062
+14,0.20176862180233002,597.4658813476562,605.9008178710938,0.294156938791275,711.6719970703125,723.9674072265625
+15,0.19844871759414673,596.2552490234375,604.551025390625,0.27007120847702026,710.80419921875,722.0928344726562
+16,0.19169431924819946,596.07421875,604.08642578125,0.2616179883480072,710.8001708984375,721.7366333007812
+17,0.18838432431221008,594.647705078125,602.5237426757812,0.2817229926586151,712.16162109375,723.937744140625
+18,0.18212516605854034,593.7442016601562,601.3579711914062,0.2628501355648041,713.4052734375,724.3924560546875

fold_1/model.bias_scaled.fold_1.ENCSR923VTG.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e50b405bf015b2ddd8df052ebf24a51c95bd9d9a5be71a30b18a0f3f970aec5f
+size 2691928

fold_1/model.bias_scaled.fold_1.ENCSR923VTG.tar

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:335439a0e703650a13562d07d4f2421fc3bc17cd21759843fa46ff99130d0534
+size 1198080

fold_1/model.chrombpnet.fold_1.ENCSR923VTG.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d00eecd3013df1817be2f5895d63f295393b9e26b35dac748acf68f21d6107a4
+size 77538896

fold_1/model.chrombpnet.fold_1.ENCSR923VTG.tar

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:160ddaad696f76897ee38b698bf2670ad51fc8f9d0edd3995009eec2da46489b
+size 27535360

fold_1/model.chrombpnet_nobias.fold_1.ENCSR923VTG.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e0a9defa743fbb88bda0a9132873a9f0e96d8b5f275f994eed2e546c2622e17
+size 25582648

fold_1/model.chrombpnet_nobias.fold_1.ENCSR923VTG.tar

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2eb1cd085afcad746a471f7e58aae8ef71ae280239d486541c7287b28dfd2aeb
+size 26060800

fold_2/logs.models.fold_2.ENCSR923VTG/logfile.modelling.fold_2.ENCSR923VTG.args.json

@@ -0,0 +1,50 @@
+{
+    "cmd": "pipeline",
+    "genome": "/oak/stanford/groups/akundaje/ziwei75/atac_seq_pipeline/hg38/GRCh38_no_alt_analysis_set_GCA_000001405.15.fasta",
+    "chrom_sizes": "/oak/stanford/groups/akundaje/ziwei75/atac_seq_pipeline/hg38/GRCh38_EBV.chrom.sizes.tsv",
+    "input_bam_file": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/preprocessing/bigWigs/ENCSR923VTG.bigWig",
+    "input_fragment_file": null,
+    "input_tagalign_file": null,
+    "output_dir": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_2/model_outputs",
+    "data_type": "ATAC",
+    "peaks": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_2/model_outputs/auxiliary/filtered.peaks.bed",
+    "nonpeaks": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_2/model_outputs/auxiliary/filtered.nonpeaks.bed",
+    "chr_fold_path": "/oak/stanford/groups/akundaje/projects/chromatin-atlas-2022/splits/fold_2.json",
+    "outlier_threshold": 0.9999,
+    "ATAC_ref_path": null,
+    "DNASE_ref_path": null,
+    "num_samples": 10000,
+    "inputlen": 2114,
+    "outputlen": 1000,
+    "seed": 1234,
+    "epochs": 50,
+    "early_stop": 5,
+    "learning_rate": 0.001,
+    "trackables": [
+        "logcount_predictions_loss",
+        "loss",
+        "logits_profile_predictions_loss",
+        "val_logcount_predictions_loss",
+        "val_loss",
+        "val_logits_profile_predictions_loss"
+    ],
+    "architecture_from_file": "/home/users/vhecht/chrombpnet/chrombpnet/chrombpnet/training/models/chrombpnet_with_bias_model.py",
+    "file_prefix": null,
+    "html_prefix": "./",
+    "bsort": false,
+    "tmpdir": null,
+    "no_st": false,
+    "bias_model_path": "/scratch/users/vhecht/chromatin-atlas-2022/reference/ATAC/fold_0/model.bias.fold_0.ENCSR868FGK.h5",
+    "negative_sampling_ratio": 0.1,
+    "filters": 512,
+    "n_dilation_layers": 8,
+    "max_jitter": 500,
+    "batch_size": 64,
+    "output_prefix": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_2/model_outputs/models/chrombpnet",
+    "bigwig": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_2/model_outputs/auxiliary/data_unstranded.bw",
+    "plus_shift": null,
+    "minus_shift": null,
+    "chr": "chr22",
+    "pwm_width": 24,
+    "params": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_2/model_outputs/logs/chrombpnet_model_params.tsv"
+}

fold_2/logs.models.fold_2.ENCSR923VTG/logfile.modelling.fold_2.ENCSR923VTG.bias_formatting.stdout.txt

@@ -0,0 +1 @@
+Converting /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_2/model_outputs/models/bias_model_scaled.h5 to /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_2/new_model_format/bias_model_scaled.tar with get_new_tf_model_format.py

fold_2/logs.models.fold_2.ENCSR923VTG/logfile.modelling.fold_2.ENCSR923VTG.chrombpnet_data_params.tsv

@@ -0,0 +1,3 @@
+counts_sum_min_thresh	0.0
+counts_sum_max_thresh	10993.86
+trainings_pts_post_thresh	199606

fold_2/logs.models.fold_2.ENCSR923VTG/logfile.modelling.fold_2.ENCSR923VTG.chrombpnet_formatting.stdout.txt

@@ -0,0 +1 @@
+Converting /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_2/model_outputs/models/chrombpnet.h5 to /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_2/new_model_format/chrombpnet.tar with get_new_tf_model_format.py

fold_2/logs.models.fold_2.ENCSR923VTG/logfile.modelling.fold_2.ENCSR923VTG.chrombpnet_model_params.tsv

@@ -0,0 +1,9 @@
+counts_loss_weight	41.7
+filters	512
+n_dil_layers	8
+bias_model_path	/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_2/model_outputs/models/bias_model_scaled.h5
+inputlen	2114
+outputlen	1000
+max_jitter	500
+chr_fold_path	/oak/stanford/groups/akundaje/projects/chromatin-atlas-2022/splits/fold_2.json
+negative_sampling_ratio	0.1

fold_2/logs.models.fold_2.ENCSR923VTG/logfile.modelling.fold_2.ENCSR923VTG.chrombpnet_no_bias_formatting.stdout.txt

@@ -0,0 +1 @@
+Converting /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_2/model_outputs/models/chrombpnet_nobias.h5 to /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_2/new_model_format/chrombpnet_nobias.tar with get_new_tf_model_format.py

fold_2/logs.models.fold_2.ENCSR923VTG/logfile.modelling.fold_2.ENCSR923VTG.epoch_loss.csv

@@ -0,0 +1,13 @@
+epoch,logcount_predictions_loss,logits_profile_predictions_loss,loss,val_logcount_predictions_loss,val_logits_profile_predictions_loss,val_loss
+0,1.3118106126785278,682.1867065429688,736.8901977539062,0.3906746804714203,680.50439453125,696.7952880859375
+1,0.3619520962238312,649.9669189453125,665.0601806640625,0.3353489339351654,671.4364013671875,685.4207153320312
+2,0.32774215936660767,639.5938110351562,653.2600708007812,0.29775118827819824,666.1229248046875,678.5390625
+3,0.30349987745285034,631.4556274414062,644.1114501953125,0.3059215843677521,663.8673706054688,676.6243896484375
+4,0.28606757521629333,624.5267944335938,636.4556274414062,0.2750419080257416,662.230712890625,673.699951171875
+5,0.2715533375740051,619.8309326171875,631.1547241210938,0.2818855941295624,663.3953247070312,675.1503295898438
+6,0.2610083818435669,616.3192138671875,627.2032470703125,0.27229300141334534,660.5955200195312,671.949951171875
+7,0.24622994661331177,613.2310791015625,623.4993896484375,0.2693490982055664,662.9412231445312,674.173095703125
+8,0.2408018261194229,611.461669921875,621.5027465820312,0.2682875394821167,661.782958984375,672.9705200195312
+9,0.2323518991470337,609.324951171875,619.014404296875,0.29649823904037476,662.1249389648438,674.4890747070312
+10,0.22551102936267853,607.4871215820312,616.8911743164062,0.26845020055770874,662.4337768554688,673.6281127929688
+11,0.21938879787921906,606.2813720703125,615.4301147460938,0.2739913761615753,662.6041870117188,674.029541015625

fold_2/model.bias_scaled.fold_2.ENCSR923VTG.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f1dc641292ba939f1f292569fab6f5c70609db19ea9e7878706413141911c4e3
+size 2691928

fold_2/model.bias_scaled.fold_2.ENCSR923VTG.tar

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e88193c008d67f29d606004a4572806dd4a03876e1022ea90dd0cf7522ae36f5
+size 1198080

fold_2/model.chrombpnet.fold_2.ENCSR923VTG.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3acde471d628defce5745d23498139dc471bf4fb1b0d0edc1b17601513ec59c3
+size 77538896

fold_2/model.chrombpnet.fold_2.ENCSR923VTG.tar

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a2516689a76aac0f4f5938a9e8940bcb079907406218e66df0bcf132593cd7c
+size 27535360

fold_2/model.chrombpnet_nobias.fold_2.ENCSR923VTG.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2240f0a9fbd4af86379cf45ad6e30cc340cc35ecd20fd8b0a645575a7ce361d5
+size 25582648

fold_2/model.chrombpnet_nobias.fold_2.ENCSR923VTG.tar

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:822c09136b2ae54ce785b059c4bd8e111a81482bdca323a57600e461705fc7b3
+size 26060800

fold_3/logs.models.fold_3.ENCSR923VTG/logfile.modelling.fold_3.ENCSR923VTG.args.json

@@ -0,0 +1,50 @@
+{
+    "cmd": "pipeline",
+    "genome": "/oak/stanford/groups/akundaje/ziwei75/atac_seq_pipeline/hg38/GRCh38_no_alt_analysis_set_GCA_000001405.15.fasta",
+    "chrom_sizes": "/oak/stanford/groups/akundaje/ziwei75/atac_seq_pipeline/hg38/GRCh38_EBV.chrom.sizes.tsv",
+    "input_bam_file": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/preprocessing/bigWigs/ENCSR923VTG.bigWig",
+    "input_fragment_file": null,
+    "input_tagalign_file": null,
+    "output_dir": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_3/model_outputs",
+    "data_type": "ATAC",
+    "peaks": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_3/model_outputs/auxiliary/filtered.peaks.bed",
+    "nonpeaks": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_3/model_outputs/auxiliary/filtered.nonpeaks.bed",
+    "chr_fold_path": "/oak/stanford/groups/akundaje/projects/chromatin-atlas-2022/splits/fold_3.json",
+    "outlier_threshold": 0.9999,
+    "ATAC_ref_path": null,
+    "DNASE_ref_path": null,
+    "num_samples": 10000,
+    "inputlen": 2114,
+    "outputlen": 1000,
+    "seed": 1234,
+    "epochs": 50,
+    "early_stop": 5,
+    "learning_rate": 0.001,
+    "trackables": [
+        "logcount_predictions_loss",
+        "loss",
+        "logits_profile_predictions_loss",
+        "val_logcount_predictions_loss",
+        "val_loss",
+        "val_logits_profile_predictions_loss"
+    ],
+    "architecture_from_file": "/home/users/vhecht/chrombpnet/chrombpnet/chrombpnet/training/models/chrombpnet_with_bias_model.py",
+    "file_prefix": null,
+    "html_prefix": "./",
+    "bsort": false,
+    "tmpdir": null,
+    "no_st": false,
+    "bias_model_path": "/scratch/users/vhecht/chromatin-atlas-2022/reference/ATAC/fold_0/model.bias.fold_0.ENCSR868FGK.h5",
+    "negative_sampling_ratio": 0.1,
+    "filters": 512,
+    "n_dilation_layers": 8,
+    "max_jitter": 500,
+    "batch_size": 64,
+    "output_prefix": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_3/model_outputs/models/chrombpnet",
+    "bigwig": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_3/model_outputs/auxiliary/data_unstranded.bw",
+    "plus_shift": null,
+    "minus_shift": null,
+    "chr": "chr6",
+    "pwm_width": 24,
+    "params": "/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_3/model_outputs/logs/chrombpnet_model_params.tsv"
+}

fold_3/logs.models.fold_3.ENCSR923VTG/logfile.modelling.fold_3.ENCSR923VTG.bias_formatting.stdout.txt

@@ -0,0 +1 @@
+Converting /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_3/model_outputs/models/bias_model_scaled.h5 to /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_3/new_model_format/bias_model_scaled.tar with get_new_tf_model_format.py

fold_3/logs.models.fold_3.ENCSR923VTG/logfile.modelling.fold_3.ENCSR923VTG.chrombpnet_data_params.tsv

@@ -0,0 +1,3 @@
+counts_sum_min_thresh	0.0
+counts_sum_max_thresh	10894.27
+trainings_pts_post_thresh	192001

fold_3/logs.models.fold_3.ENCSR923VTG/logfile.modelling.fold_3.ENCSR923VTG.chrombpnet_formatting.stdout.txt

@@ -0,0 +1 @@
+Converting /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_3/model_outputs/models/chrombpnet.h5 to /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_3/new_model_format/chrombpnet.tar with get_new_tf_model_format.py

fold_3/logs.models.fold_3.ENCSR923VTG/logfile.modelling.fold_3.ENCSR923VTG.chrombpnet_model_params.tsv

@@ -0,0 +1,9 @@
+counts_loss_weight	41.9
+filters	512
+n_dil_layers	8
+bias_model_path	/scratch/users/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_3/model_outputs/models/bias_model_scaled.h5
+inputlen	2114
+outputlen	1000
+max_jitter	500
+chr_fold_path	/oak/stanford/groups/akundaje/projects/chromatin-atlas-2022/splits/fold_3.json
+negative_sampling_ratio	0.1

fold_3/logs.models.fold_3.ENCSR923VTG/logfile.modelling.fold_3.ENCSR923VTG.chrombpnet_no_bias_formatting.stdout.txt

@@ -0,0 +1 @@
+Converting /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_3/model_outputs/models/chrombpnet_nobias.h5 to /oak/stanford/groups/akundaje/vhecht/chromatin-atlas-2022/ATAC/ENCSR923VTG/fold_3/new_model_format/chrombpnet_nobias.tar with get_new_tf_model_format.py