broadinstitute · gwaybio · Apr 1, 2020 · Mar 30, 2020 · Mar 30, 2020 · Mar 30, 2020
diff --git a/1.generate-profiles/4.load-example-cell-painting-image.ipynb b/1.generate-profiles/4.load-example-cell-painting-image.ipynb
diff --git a/1.generate-profiles/figures/example_images/color_channel.png b/1.generate-profiles/figures/example_images/color_channel.png
diff --git a/1.generate-profiles/figures/example_images/color_merged.png b/1.generate-profiles/figures/example_images/color_merged.png
diff --git a/1.generate-profiles/figures/example_images/greyscale_channel.png b/1.generate-profiles/figures/example_images/greyscale_channel.png
diff --git a/1.generate-profiles/scripts/nbconverted/4.load-example-cell-painting-image.py b/1.generate-profiles/scripts/nbconverted/4.load-example-cell-painting-image.py
@@ -0,0 +1,128 @@
+#!/usr/bin/env python
+# coding: utf-8
+
+# # Load an Example Image for Approach Figure
+# 
+# Here, I use `skimage` to load, process, and output example images across channels and an example composite image with merged channels.
+# 
+# Note that much of this code was adapted from https://github.com/jr0th/segmentation/blob/master/visualization/CellArt.ipynb and with insights from Marzieh Haghighi.
+
+# In[1]:
+
+
+import os
+import sys
+import numpy as np
+import pandas as pd
+import matplotlib.pyplot as plt
+
+import skimage.io
+import skimage.exposure
+
+from pycytominer.aggregate import AggregateProfiles
+
+
+# In[2]:
+
+
+sys.path.append("../scripts")
+from grabPictures import grabPicture
+
+
+# ## Set File Paths
+
+# In[3]:
+
+
+bucket_path = "/home/ubuntu/bucket/projects/2015_07_01_Cell_Health_Vazquez_Cancer_Broad/"
+image_path = os.path.join(bucket_path, "CRISPR_PILOT_B1/images/")
+sqlite_path = os.path.join(bucket_path, "workspace/backend/CRISPR_PILOT_B1/")
+metadata_path = os.path.join("data", "metadata")
+
+
+# In[4]:
+
+
+plate = "SQ00014613"
+sqlite_dir = os.path.join(sqlite_path, plate)
+sqlite_file = "sqlite:///{}/{}.sqlite".format(sqlite_dir, plate)
+sqlite_file
+
+
+# In[5]:
+
+
+barcode_file = os.path.join(metadata_path, "barcode_platemap.csv")
+platemap_name = pd.read_csv(barcode_file).query("Assay_Plate_Barcode == @plate").Plate_Map_Name.values[0]
+platemap_path = os.path.join(metadata_path, "platemap", "{}.csv".format(platemap_name))
+
+
+# ## Process Data and Load Example Images
+
+# In[6]:
+
+
+gene_name = "EMPTY"
+pert_name = "EMPTY"
+site = 1
+
+
+# In[7]:
+
+
+example_image = grabPicture(
+    bucket_path=bucket_path,
+    image_path=image_path,
+    sqlite_path=sqlite_path,
+    metadata_path=metadata_path,
+    plate=plate,
+    aggregate_strata_cols=["Image_Metadata_Plate", "Image_Metadata_Well"],
+    gene_name=gene_name,
+    pert_name=pert_name,
+    site=site,
+)
+
+
+# In[8]:
+
+
+example_image.load_image_table(get_well=True)
+
+
+# In[9]:
+
+
+# Prep images with a narrow crop
+example_image.prep_images(low_prop=0.1, high_prop=0.3)
+
+
+# ## Output an Example Image
+
+# In[10]:
+
+
+example_image.plot_images(cropped=True, color=False)
+
+output_file = os.path.join("figures", "example_images", "greyscale_channel.png")
+plt.subplots_adjust(top=0.8)
+plt.savefig(output_file, pad_inches=0.2, dpi=500)
+
+
+# In[11]:
+
+
+example_image.plot_images(cropped=True, color=True)
+
+output_file = os.path.join("figures", "example_images", "color_channel.png")
+plt.subplots_adjust(top=0.8)
+plt.savefig(output_file, pad_inches=0.2, dpi=500)
+
+
+# In[12]:
+
+
+example_image.plot_combined_image()
+
+output_file = os.path.join("figures", "example_images", "color_merged.png")
+plt.savefig(output_file, dpi=500)
+
diff --git a/3.train/4.visualize-regression-performance.ipynb b/3.train/4.visualize-regression-performance.ipynb
diff --git a/3.train/7.visualize-cell-line-performance.ipynb b/3.train/7.visualize-cell-line-performance.ipynb
@@ -557,12 +557,8 @@
     "    xlab(\"CRISPR Efficiency\") + \n",
     "    scale_color_manual(\n",
     "        name = \"Cell Line\",\n",
-    "        labels = c(\"A549\" = \"A549\",\n",
-    "                   \"ES2\" = \"ES2\",\n",
-    "                   \"HCC44\" = \"HCC44\"),\n",
-    "        values = c(\"A549\" = \"#7fc97f\",\n",
-    "                   \"ES2\" = \"#beaed4\",\n",
-    "                   \"HCC44\" = \"#fdc086\")\n",
+    "        labels = cell_line_labels,\n",
+    "        values = cell_line_colors\n",
     "    ) +\n",
     "    coord_fixed() +\n",
     "    theme_bw() +\n",

diff --git a/...arget_performance/regression/median/cc_all_high_h2ax_raw_performance_median.png b/...arget_performance/regression/median/cc_all_high_h2ax_raw_performance_median.png
diff --git a/...ession/median/cc_all_large_notround_polynuclear_mean_raw_performance_median.png b/...ession/median/cc_all_large_notround_polynuclear_mean_raw_performance_median.png
diff --git a/.../regression/median/cc_all_large_round_polyploid_mean_raw_performance_median.png b/.../regression/median/cc_all_large_round_polyploid_mean_raw_performance_median.png
diff --git a/...arget_performance/regression/median/cc_all_n_objects_raw_performance_median.png b/...arget_performance/regression/median/cc_all_n_objects_raw_performance_median.png
diff --git a/...rformance/regression/median/cc_all_n_spots_h2ax_mean_raw_performance_median.png b/...rformance/regression/median/cc_all_n_spots_h2ax_mean_raw_performance_median.png
diff --git a/...ion/median/cc_all_n_spots_h2ax_per_nucleus_area_mean_raw_performance_median.png b/...ion/median/cc_all_n_spots_h2ax_per_nucleus_area_mean_raw_performance_median.png
diff --git a/...rformance/regression/median/cc_all_nucleus_area_mean_raw_performance_median.png b/...rformance/regression/median/cc_all_nucleus_area_mean_raw_performance_median.png
diff --git a/...ance/regression/median/cc_all_nucleus_roundness_mean_raw_performance_median.png b/...ance/regression/median/cc_all_nucleus_roundness_mean_raw_performance_median.png
diff --git a/...et_performance/regression/median/cc_cc_early_mitosis_raw_performance_median.png b/...et_performance/regression/median/cc_cc_early_mitosis_raw_performance_median.png
diff --git a/...vidual_target_performance/regression/median/cc_cc_g1_raw_performance_median.png b/...vidual_target_performance/regression/median/cc_cc_g1_raw_performance_median.png
diff --git a/...vidual_target_performance/regression/median/cc_cc_g2_raw_performance_median.png b/...vidual_target_performance/regression/median/cc_cc_g2_raw_performance_median.png
diff --git a/...target_performance/regression/median/cc_cc_high_h2ax_raw_performance_median.png b/...target_performance/regression/median/cc_cc_high_h2ax_raw_performance_median.png
diff --git a/...get_performance/regression/median/cc_cc_late_mitosis_raw_performance_median.png b/...get_performance/regression/median/cc_cc_late_mitosis_raw_performance_median.png
diff --git a/...l_target_performance/regression/median/cc_cc_mitosis_raw_performance_median.png b/...l_target_performance/regression/median/cc_cc_mitosis_raw_performance_median.png
diff --git a/...target_performance/regression/median/cc_cc_n_objects_raw_performance_median.png b/...target_performance/regression/median/cc_cc_n_objects_raw_performance_median.png
diff --git a/...erformance/regression/median/cc_cc_n_spots_h2ax_mean_raw_performance_median.png b/...erformance/regression/median/cc_cc_n_spots_h2ax_mean_raw_performance_median.png
diff --git a/...sion/median/cc_cc_n_spots_h2ax_per_nucleus_area_mean_raw_performance_median.png b/...sion/median/cc_cc_n_spots_h2ax_per_nucleus_area_mean_raw_performance_median.png
diff --git a/...ividual_target_performance/regression/median/cc_cc_s_raw_performance_median.png b/...ividual_target_performance/regression/median/cc_cc_s_raw_performance_median.png
diff --git a/...ormance/regression/median/cc_early_mitosis_n_objects_raw_performance_median.png b/...ormance/regression/median/cc_early_mitosis_n_objects_raw_performance_median.png
diff --git a/.../cc_early_mitosis_n_spots_h2ax_per_nucleus_area_mean_raw_performance_median.png b/.../cc_early_mitosis_n_spots_h2ax_per_nucleus_area_mean_raw_performance_median.png
diff --git a/...target_performance/regression/median/cc_g1_n_objects_raw_performance_median.png b/...target_performance/regression/median/cc_g1_n_objects_raw_performance_median.png
diff --git a/...erformance/regression/median/cc_g1_n_spots_h2ax_mean_raw_performance_median.png b/...erformance/regression/median/cc_g1_n_spots_h2ax_mean_raw_performance_median.png
diff --git a/...sion/median/cc_g1_n_spots_h2ax_per_nucleus_area_mean_raw_performance_median.png b/...sion/median/cc_g1_n_spots_h2ax_per_nucleus_area_mean_raw_performance_median.png
diff --git a/...et_performance/regression/median/cc_g1_plus_g2_count_raw_performance_median.png b/...et_performance/regression/median/cc_g1_plus_g2_count_raw_performance_median.png
diff --git a/...target_performance/regression/median/cc_g2_n_objects_raw_performance_median.png b/...target_performance/regression/median/cc_g2_n_objects_raw_performance_median.png
diff --git a/...sion/median/cc_g2_n_spots_h2ax_per_nucleus_area_mean_raw_performance_median.png b/...sion/median/cc_g2_n_spots_h2ax_per_nucleus_area_mean_raw_performance_median.png
diff --git a/...performance/regression/median/cc_g2_plus_all_m_count_raw_performance_median.png b/...performance/regression/median/cc_g2_plus_all_m_count_raw_performance_median.png
diff --git a/...erformance/regression/median/cc_infection_percentage_raw_performance_median.png b/...erformance/regression/median/cc_infection_percentage_raw_performance_median.png
diff --git a/...formance/regression/median/cc_late_mitosis_n_objects_raw_performance_median.png b/...formance/regression/median/cc_late_mitosis_n_objects_raw_performance_median.png
diff --git a/...t_performance/regression/median/cc_mitosis_n_objects_raw_performance_median.png b/...t_performance/regression/median/cc_mitosis_n_objects_raw_performance_median.png
diff --git a/...mance/regression/median/cc_mitosis_n_spots_h2ax_mean_raw_performance_median.png b/...mance/regression/median/cc_mitosis_n_spots_h2ax_mean_raw_performance_median.png
diff --git a/...rformance/regression/median/cc_polynuclear_n_objects_raw_performance_median.png b/...rformance/regression/median/cc_polynuclear_n_objects_raw_performance_median.png
diff --git a/...performance/regression/median/cc_polyploid_high_h2ax_raw_performance_median.png b/...performance/regression/median/cc_polyploid_high_h2ax_raw_performance_median.png
diff --git a/...performance/regression/median/cc_polyploid_n_objects_raw_performance_median.png b/...performance/regression/median/cc_polyploid_n_objects_raw_performance_median.png
diff --git a/...nce/regression/median/cc_polyploid_n_spots_h2ax_mean_raw_performance_median.png b/...nce/regression/median/cc_polyploid_n_spots_h2ax_mean_raw_performance_median.png
diff --git a/..._target_performance/regression/median/cc_s_high_h2ax_raw_performance_median.png b/..._target_performance/regression/median/cc_s_high_h2ax_raw_performance_median.png
diff --git a/...e/regression/median/cc_s_intensity_nucleus_area_mean_raw_performance_median.png b/...e/regression/median/cc_s_intensity_nucleus_area_mean_raw_performance_median.png
diff --git a/...ce/regression/median/cc_s_intensity_nucleus_area_sum_raw_performance_median.png b/...ce/regression/median/cc_s_intensity_nucleus_area_sum_raw_performance_median.png
diff --git a/..._target_performance/regression/median/cc_s_n_objects_raw_performance_median.png b/..._target_performance/regression/median/cc_s_n_objects_raw_performance_median.png
diff --git a/...erformance/regression/median/vb_infection_percentage_raw_performance_median.png b/...erformance/regression/median/vb_infection_percentage_raw_performance_median.png
diff --git a/...rget_performance/regression/median/vb_live_cell_area_raw_performance_median.png b/...rget_performance/regression/median/vb_live_cell_area_raw_performance_median.png
diff --git a/...performance/regression/median/vb_live_cell_roundness_raw_performance_median.png b/...performance/regression/median/vb_live_cell_roundness_raw_performance_median.png
diff --git a/...formance/regression/median/vb_live_cell_width_length_raw_performance_median.png b/...formance/regression/median/vb_live_cell_width_length_raw_performance_median.png
diff --git a/...rget_performance/regression/median/vb_num_live_cells_raw_performance_median.png b/...rget_performance/regression/median/vb_num_live_cells_raw_performance_median.png
diff --git a/...rformance/regression/median/vb_percent_all_apoptosis_raw_performance_median.png b/...rformance/regression/median/vb_percent_all_apoptosis_raw_performance_median.png
diff --git a/...t_performance/regression/median/vb_percent_dead_only_raw_performance_median.png b/...t_performance/regression/median/vb_percent_dead_only_raw_performance_median.png
diff --git a/...target_performance/regression/median/vb_percent_dead_raw_performance_median.png b/...target_performance/regression/median/vb_percent_dead_raw_performance_median.png
diff --git a/...formance/regression/median/vb_percent_late_apoptosis_raw_performance_median.png b/...formance/regression/median/vb_percent_late_apoptosis_raw_performance_median.png
diff --git a/...target_performance/regression/median/vb_percent_live_raw_performance_median.png b/...target_performance/regression/median/vb_percent_live_raw_performance_median.png
diff --git a/...arget_performance/regression/median/vb_ros_back_mean_raw_performance_median.png b/...arget_performance/regression/median/vb_ros_back_mean_raw_performance_median.png
diff --git a/...ual_target_performance/regression/median/vb_ros_mean_raw_performance_median.png b/...ual_target_performance/regression/median/vb_ros_mean_raw_performance_median.png
diff --git a/...regression/modz/cc_all_large_notround_polynuclear_mean_raw_performance_modz.png b/...regression/modz/cc_all_large_notround_polynuclear_mean_raw_performance_modz.png
diff --git a/...ance/regression/modz/cc_all_large_round_polyploid_mean_raw_performance_modz.png b/...ance/regression/modz/cc_all_large_round_polyploid_mean_raw_performance_modz.png
diff --git a/...al_target_performance/regression/modz/cc_all_n_objects_raw_performance_modz.png b/...al_target_performance/regression/modz/cc_all_n_objects_raw_performance_modz.png
diff --git a/...t_performance/regression/modz/cc_all_n_spots_h2ax_mean_raw_performance_modz.png b/...t_performance/regression/modz/cc_all_n_spots_h2ax_mean_raw_performance_modz.png
diff --git a/...ression/modz/cc_all_n_spots_h2ax_per_nucleus_area_mean_raw_performance_modz.png b/...ression/modz/cc_all_n_spots_h2ax_per_nucleus_area_mean_raw_performance_modz.png
diff --git a/...t_performance/regression/modz/cc_all_nucleus_area_mean_raw_performance_modz.png b/...t_performance/regression/modz/cc_all_nucleus_area_mean_raw_performance_modz.png
diff --git a/...formance/regression/modz/cc_all_nucleus_roundness_mean_raw_performance_modz.png b/...formance/regression/modz/cc_all_nucleus_roundness_mean_raw_performance_modz.png
diff --git a/...individual_target_performance/regression/modz/cc_cc_g1_raw_performance_modz.png b/...individual_target_performance/regression/modz/cc_cc_g1_raw_performance_modz.png
diff --git a/...individual_target_performance/regression/modz/cc_cc_g2_raw_performance_modz.png b/...individual_target_performance/regression/modz/cc_cc_g2_raw_performance_modz.png
diff --git a/...ual_target_performance/regression/modz/cc_cc_high_h2ax_raw_performance_modz.png b/...ual_target_performance/regression/modz/cc_cc_high_h2ax_raw_performance_modz.png
diff --git a/..._target_performance/regression/modz/cc_cc_late_mitosis_raw_performance_modz.png b/..._target_performance/regression/modz/cc_cc_late_mitosis_raw_performance_modz.png
diff --git a/...idual_target_performance/regression/modz/cc_cc_mitosis_raw_performance_modz.png b/...idual_target_performance/regression/modz/cc_cc_mitosis_raw_performance_modz.png
diff --git a/...ual_target_performance/regression/modz/cc_cc_n_objects_raw_performance_modz.png b/...ual_target_performance/regression/modz/cc_cc_n_objects_raw_performance_modz.png
diff --git a/...et_performance/regression/modz/cc_cc_n_spots_h2ax_mean_raw_performance_modz.png b/...et_performance/regression/modz/cc_cc_n_spots_h2ax_mean_raw_performance_modz.png
diff --git a/...gression/modz/cc_cc_n_spots_h2ax_per_nucleus_area_mean_raw_performance_modz.png b/...gression/modz/cc_cc_n_spots_h2ax_per_nucleus_area_mean_raw_performance_modz.png
diff --git a/.../individual_target_performance/regression/modz/cc_cc_s_raw_performance_modz.png b/.../individual_target_performance/regression/modz/cc_cc_s_raw_performance_modz.png
diff --git a/...performance/regression/modz/cc_early_mitosis_high_h2ax_raw_performance_modz.png b/...performance/regression/modz/cc_early_mitosis_high_h2ax_raw_performance_modz.png
diff --git a/...performance/regression/modz/cc_early_mitosis_n_objects_raw_performance_modz.png b/...performance/regression/modz/cc_early_mitosis_n_objects_raw_performance_modz.png
diff --git a/...nce/regression/modz/cc_early_mitosis_n_spots_h2ax_mean_raw_performance_modz.png b/...nce/regression/modz/cc_early_mitosis_n_spots_h2ax_mean_raw_performance_modz.png
diff --git a/...ual_target_performance/regression/modz/cc_g1_n_objects_raw_performance_modz.png b/...ual_target_performance/regression/modz/cc_g1_n_objects_raw_performance_modz.png
diff --git a/...et_performance/regression/modz/cc_g1_n_spots_h2ax_mean_raw_performance_modz.png b/...et_performance/regression/modz/cc_g1_n_spots_h2ax_mean_raw_performance_modz.png
diff --git a/...gression/modz/cc_g1_n_spots_h2ax_per_nucleus_area_mean_raw_performance_modz.png b/...gression/modz/cc_g1_n_spots_h2ax_per_nucleus_area_mean_raw_performance_modz.png
diff --git a/...target_performance/regression/modz/cc_g1_plus_g2_count_raw_performance_modz.png b/...target_performance/regression/modz/cc_g1_plus_g2_count_raw_performance_modz.png
diff --git a/...ual_target_performance/regression/modz/cc_g2_n_objects_raw_performance_modz.png b/...ual_target_performance/regression/modz/cc_g2_n_objects_raw_performance_modz.png
diff --git a/...et_performance/regression/modz/cc_g2_n_spots_h2ax_mean_raw_performance_modz.png b/...et_performance/regression/modz/cc_g2_n_spots_h2ax_mean_raw_performance_modz.png
diff --git a/...get_performance/regression/modz/cc_g2_plus_all_m_count_raw_performance_modz.png b/...get_performance/regression/modz/cc_g2_plus_all_m_count_raw_performance_modz.png
diff --git a/...et_performance/regression/modz/cc_infection_percentage_raw_performance_modz.png b/...et_performance/regression/modz/cc_infection_percentage_raw_performance_modz.png
diff --git a/..._performance/regression/modz/cc_late_mitosis_n_objects_raw_performance_modz.png b/..._performance/regression/modz/cc_late_mitosis_n_objects_raw_performance_modz.png
diff --git a/...odz/cc_late_mitosis_n_spots_h2ax_per_nucleus_area_mean_raw_performance_modz.png b/...odz/cc_late_mitosis_n_spots_h2ax_per_nucleus_area_mean_raw_performance_modz.png
diff --git a/...arget_performance/regression/modz/cc_mitosis_n_objects_raw_performance_modz.png b/...arget_performance/regression/modz/cc_mitosis_n_objects_raw_performance_modz.png
diff --git a/...t_performance/regression/modz/cc_polynuclear_high_h2ax_raw_performance_modz.png b/...t_performance/regression/modz/cc_polynuclear_high_h2ax_raw_performance_modz.png
diff --git a/...t_performance/regression/modz/cc_polynuclear_n_objects_raw_performance_modz.png b/...t_performance/regression/modz/cc_polynuclear_n_objects_raw_performance_modz.png
diff --git a/...modz/cc_polynuclear_n_spots_h2ax_per_nucleus_area_mean_raw_performance_modz.png b/...modz/cc_polynuclear_n_spots_h2ax_per_nucleus_area_mean_raw_performance_modz.png
diff --git a/...get_performance/regression/modz/cc_polyploid_high_h2ax_raw_performance_modz.png b/...get_performance/regression/modz/cc_polyploid_high_h2ax_raw_performance_modz.png
diff --git a/...get_performance/regression/modz/cc_polyploid_n_objects_raw_performance_modz.png b/...get_performance/regression/modz/cc_polyploid_n_objects_raw_performance_modz.png
diff --git a/...n/modz/cc_polyploid_n_spots_h2ax_per_nucleus_area_mean_raw_performance_modz.png b/...n/modz/cc_polyploid_n_spots_h2ax_per_nucleus_area_mean_raw_performance_modz.png
diff --git a/...dual_target_performance/regression/modz/cc_s_high_h2ax_raw_performance_modz.png b/...dual_target_performance/regression/modz/cc_s_high_h2ax_raw_performance_modz.png
diff --git a/...mance/regression/modz/cc_s_intensity_nucleus_area_mean_raw_performance_modz.png b/...mance/regression/modz/cc_s_intensity_nucleus_area_mean_raw_performance_modz.png
diff --git a/...rmance/regression/modz/cc_s_intensity_nucleus_area_sum_raw_performance_modz.png b/...rmance/regression/modz/cc_s_intensity_nucleus_area_sum_raw_performance_modz.png
diff --git a/...dual_target_performance/regression/modz/cc_s_n_objects_raw_performance_modz.png b/...dual_target_performance/regression/modz/cc_s_n_objects_raw_performance_modz.png
diff --git a/...get_performance/regression/modz/cc_s_n_spots_h2ax_mean_raw_performance_modz.png b/...get_performance/regression/modz/cc_s_n_spots_h2ax_mean_raw_performance_modz.png
diff --git a/...egression/modz/cc_s_n_spots_h2ax_per_nucleus_area_mean_raw_performance_modz.png b/...egression/modz/cc_s_n_spots_h2ax_per_nucleus_area_mean_raw_performance_modz.png
diff --git a/...et_performance/regression/modz/vb_infection_percentage_raw_performance_modz.png b/...et_performance/regression/modz/vb_infection_percentage_raw_performance_modz.png
diff --git a/...l_target_performance/regression/modz/vb_live_cell_area_raw_performance_modz.png b/...l_target_performance/regression/modz/vb_live_cell_area_raw_performance_modz.png
diff --git a/...get_performance/regression/modz/vb_live_cell_roundness_raw_performance_modz.png b/...get_performance/regression/modz/vb_live_cell_roundness_raw_performance_modz.png
diff --git a/..._performance/regression/modz/vb_live_cell_width_length_raw_performance_modz.png b/..._performance/regression/modz/vb_live_cell_width_length_raw_performance_modz.png
diff --git a/...l_target_performance/regression/modz/vb_num_live_cells_raw_performance_modz.png b/...l_target_performance/regression/modz/vb_num_live_cells_raw_performance_modz.png
diff --git a/...t_performance/regression/modz/vb_percent_all_apoptosis_raw_performance_modz.png b/...t_performance/regression/modz/vb_percent_all_apoptosis_raw_performance_modz.png
diff --git a/...rformance/regression/modz/vb_percent_caspase_dead_only_raw_performance_modz.png b/...rformance/regression/modz/vb_percent_caspase_dead_only_raw_performance_modz.png
diff --git a/...arget_performance/regression/modz/vb_percent_dead_only_raw_performance_modz.png b/...arget_performance/regression/modz/vb_percent_dead_only_raw_performance_modz.png
diff --git a/...ual_target_performance/regression/modz/vb_percent_dead_raw_performance_modz.png b/...ual_target_performance/regression/modz/vb_percent_dead_raw_performance_modz.png
diff --git a/..._performance/regression/modz/vb_percent_late_apoptosis_raw_performance_modz.png b/..._performance/regression/modz/vb_percent_late_apoptosis_raw_performance_modz.png
diff --git a/...ual_target_performance/regression/modz/vb_percent_live_raw_performance_modz.png b/...ual_target_performance/regression/modz/vb_percent_live_raw_performance_modz.png
diff --git a/...al_target_performance/regression/modz/vb_ros_back_mean_raw_performance_modz.png b/...al_target_performance/regression/modz/vb_ros_back_mean_raw_performance_modz.png
diff --git a/...ividual_target_performance/regression/modz/vb_ros_mean_raw_performance_modz.png b/...ividual_target_performance/regression/modz/vb_ros_mean_raw_performance_modz.png
diff --git a/3.train/figures/regression/median/all_regression_performance_metrics_median.pdf b/3.train/figures/regression/median/all_regression_performance_metrics_median.pdf
diff --git a/3.train/figures/regression/median/cell_line_differences_rsquared_median.png b/3.train/figures/regression/median/cell_line_differences_rsquared_median.png
diff --git a/3.train/figures/regression/median/r_squared_comparison_scatter_median.png b/3.train/figures/regression/median/r_squared_comparison_scatter_median.png
diff --git a/3.train/figures/regression/median/regression_performance_figure_median.png b/3.train/figures/regression/median/regression_performance_figure_median.png
diff --git a/3.train/figures/regression/median/supplementary_figure_example_distributions.png b/3.train/figures/regression/median/supplementary_figure_example_distributions.png
diff --git a/3.train/figures/regression/modz/all_regression_performance_metrics_modz.pdf b/3.train/figures/regression/modz/all_regression_performance_metrics_modz.pdf
diff --git a/3.train/figures/regression/modz/cell_line_differences_rsquared_modz.png b/3.train/figures/regression/modz/cell_line_differences_rsquared_modz.png
diff --git a/3.train/figures/regression/modz/cell_line_differences_target_linked_full_modz.png b/3.train/figures/regression/modz/cell_line_differences_target_linked_full_modz.png
diff --git a/3.train/figures/regression/modz/r_squared_comparison_scatter_modz.png b/3.train/figures/regression/modz/r_squared_comparison_scatter_modz.png
diff --git a/3.train/figures/regression/modz/regression_performance_figure_modz.png b/3.train/figures/regression/modz/regression_performance_figure_modz.png
diff --git a/3.train/figures/regression/modz/supplementary_figure_example_distributions.png b/3.train/figures/regression/modz/supplementary_figure_example_distributions.png
diff --git a/3.train/scripts/assay_themes.R b/3.train/scripts/assay_themes.R
@@ -98,6 +98,18 @@ dye_labels <- c(
   "qc" = "CRISPR Efficiency"
 )
 
+cell_line_labels <- c(
+  "A549" = "A549",
+  "ES2" = "ES2",
+  "HCC44" = "HCC44"
+)
+
+cell_line_colors <- c(
+  "A549" = "#7fc97f",
+  "ES2" = "#beaed4",
+  "HCC44" = "#fdc086"
+)
+
 dye_theme <- theme(
     axis.title = element_text(size = 4),
     axis.title.x = element_text(margin = margin(0, 0, 0, 0)),

diff --git a/3.train/scripts/nbconverted/4.visualize-regression-performance.r b/3.train/scripts/nbconverted/4.visualize-regression-performance.r
@@ -293,9 +293,11 @@ rsquared_bar_cellline_gg <- ggplot(cellline_compare_regression_df,
     xlab(bquote("Test Set "~R^2~"")) +
     ylab("Cell Health Target") +
     facet_wrap(~outlier, scales="free_x") +
-    scale_color_manual(name = "Cell Line",
-                       values = c("A549" = "#A1473A", "ES2" = "#209481", "HCC44" = "#241B3B"),
-                       labels = c("A549" = "A549", "ES2" = "ES2", "HCC44" = "HCC44")) +
+    scale_color_manual(
+        name = "Cell Line",
+        labels = cell_line_labels,
+        values = cell_line_colors
+    ) +
     theme(axis.text.x = element_text(size = 8, angle = 90),
           axis.text.y = element_text(size = 5.5),
           axis.title = element_text(size = 10),
@@ -376,6 +378,7 @@ ggplot(r2_spread_df,
     ggtitle("Regression Performance") +
     theme_bw() +
     facet_wrap("~data_type") +
+    coord_fixed() +
     scale_color_manual(
         name = "Cell Health\nPhenotypes",
         values = measurement_colors,
@@ -445,6 +448,7 @@ for (good_model in c(good_example_models, bad_example_models)) {
     plot_list[[good_model]] <- ggplot(sup_fig_good_subset_df,
            aes(x = recode_target_value_true,
                y = recode_target_value_pred)) +
+        geom_smooth(method='lm', formula=y~x) +
         geom_point(aes(color = Metadata_cell_line),
                    size = 0.5,
                    alpha = 0.8) +
@@ -453,14 +457,11 @@ for (good_model in c(good_example_models, bad_example_models)) {
         theme_bw() +
         xlab("True Values") +
         ylab("Predicted Values") +
-        scale_color_manual(name = "Cell Line",
-                           labels = c("A549" = "A549",
-                                      "ES2" = "ES2",
-                                      "HCC44" = "HCC44"),
-                           values = c("A549" = "#7fc97f",
-                                      "ES2" = "#beaed4",
-                                      "HCC44" = "#fdc086")) +
-        geom_smooth(method='lm', formula=y~x) +
+        scale_color_manual(
+            name = "Cell Line",
+            labels = cell_line_labels,
+            values = cell_line_colors
+        ) +
         geom_text(data = r2_df, size = 3, aes(label = paste("R2 =", r2), x = x, y = y)) +
         theme(strip.text = element_text(size = 10),
               strip.background = element_rect(colour = "black",
@@ -580,13 +581,11 @@ for (target in unique(y_plot_df$target)) {
                 theme_bw() +
                 xlab("True Values") +
                 ylab("Predicted Values") +
-                scale_color_manual(name = "Cell Line",
-                                   labels = c("A549" = "A549",
-                                              "ES2" = "ES2",
-                                              "HCC44" = "HCC44"),
-                                   values = c("A549" = "#7fc97f",
-                                              "ES2" = "#beaed4",
-                                              "HCC44" = "#fdc086")) +
+                scale_color_manual(
+                    name = "Cell Line",
+                    labels = cell_line_labels,
+                    values = cell_line_colors
+                ) +
                 geom_smooth(method='lm', formula=y~x) +
                 theme(strip.text = element_text(size = 10),
                       strip.background = element_rect(colour = "black",

diff --git a/3.train/scripts/nbconverted/7.visualize-cell-line-performance.r b/3.train/scripts/nbconverted/7.visualize-cell-line-performance.r
@@ -172,12 +172,8 @@ ggplot(summary_df, aes(x = mean_crispr_eff, y = mean_rsquared_diff, color = Meta
     xlab("CRISPR Efficiency") + 
     scale_color_manual(
         name = "Cell Line",
-        labels = c("A549" = "A549",
-                   "ES2" = "ES2",
-                   "HCC44" = "HCC44"),
-        values = c("A549" = "#7fc97f",
-                   "ES2" = "#beaed4",
-                   "HCC44" = "#fdc086")
+        labels = cell_line_labels,
+        values = cell_line_colors
     ) +
     coord_fixed() +
     theme_bw() +

diff --git a/scripts/__init__.py b/scripts/__init__.py