RunDeepProfiler

documentation/CP-plugins-documentation/RunDeepProfiler.md

@@ -0,0 +1,65 @@
+# RunDeepProfiler
+
+## How to use RunDeepProfiler
+
+### Windows
+
+1. Follow the instructions to install CellProfiler from source: [Installing plugins with dependencies, using CellProfiler from source](using_plugins.md);
+
+2. Clone DeepProfiler repository:
+
+```
+git clone https://github.com/broadinstitute/DeepProfiler.git
+```
+
+4. Install DeepProfiler
+
+```
+cd DeepProfiler /
+pip install -e .
+```
+
+5. Install dependencies by running:
+
+```bash
+cd CellProfiler-plugins
+pip install -e .[deepprofiler]
+```
+
+## Run Example
+
+1. Use the folder with images and files available on CellProfiler-plugins > test > test_deepprofiler [link here](https://github.com/CellProfiler/CellProfiler-plugins/pull/182/commits/62874b4a28a370cea069662d3804a68b651130ec) as an example to run the test pipeline test_deepprofiler.cppipe
+
+2. Don't forget to select the config, model, and DeepProfiler directories to your local paths.
+
+
+## Using GPU
+
+If you want to use a GPU to run the model (this is recommended for speed), you'll need a compatible version of Tensorflow and a supported GPU. 
+General instructions are available at this [link](https://www.tensorflow.org/guide/gpu).
+
+1. Your GPU should be visible in Device Manager under Display Adaptors. 
+If your GPU isn't there, you likely need to install drivers.
+[Here](https://www.nvidia.com/Download/Find.aspx) is where you can find NVIDIA GPU drivers if you need to install them.
+
+
+2. To test whether the GPU is configured correctly:
+  * Run `python` on the command line (i.e., in Command Prompt or Terminal) to start an interactive session
+  * Then run the following
+  ```
+  import tensorflow as tf
+  tf.test.is_gpu_available(
+    cuda_only=False, min_cuda_compute_capability=None
+)
+  ```
+  * If this returns `True`, you're all set
+  * If this returns `False`, you likely need to install/reinstall torch. See [here](https://www.tensorflow.org/guide/gpu) for your exact command.
+  * Exit the session with `exit()` then install tensorflow if necessary.
+
+
+**NOTE**: You might get a warning like this:
+```
+W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'cudart64_110.dll'; dlerror: cudart64_110.dll not found
+2022-05-26 20:24:21.906286: I tensorflow/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+```
+If you don't have a GPU, this is not a problem. If you do, your configuration is incorrect and you need to try reinstalling drivers and the correct version of CUDA for your system.

documentation/CP-plugins-documentation/supported_plugins.md

@@ -24,4 +24,5 @@ Those plugins that do have extra documentation contain links below.
 | RunImageJScript       | RunImageJScript allows you to run any supported ImageJ script directly within CellProfiler. It is significantly more performant than RunImageJMacro, and is also less likely to leave behind temporary files. | Yes | XXXXX |
 | RunOmnipose           | RunOmnipose allows you to run Omnipose within CellProfiler. Omnipose is a general image segmentation tool that builds on Cellpose. | Yes | `omnipose` |
 | RunStarDist           | RunStarDist allows you to run StarDist within CellProfiler. StarDist is a machine-learning algorithm for object detection with star-convex shapes making it best suited for nuclei or round-ish cells. You can use pre-trained StarDist models or your custom model with this plugin. You can use a GPU with this module to dramatically increase your speed/efficiency. RunStarDist is generally faster than RunCellpose. | Yes | `stardist` |
+| [RunDeepProfiler](RunDeepProfiler.md) | RunDeepProfiler allows you to run DeepProfiler within CellProfiler. DeepProfiler can measure features from single-cells using pre-trained CNN. | Yes | `deepprofiler` |
 | VarianceTransform     | This module allows you to calculate the variance of an image, using a determined window size. It also has the option to find the optimal window size from a predetermined range to obtain the maximum variance of an image. | No | |

documentation/CP-plugins-documentation/using_plugins.md

@@ -20,6 +20,8 @@ See [Installing plugins with dependencies, using CellProfiler from source](#inst
 The second option allows you to use pre-built CellProfiler, but plugin installation is more complex.
 See [Installing plugins with dependencies, using pre-built CellProfiler](#installing-plugins-with-dependencies-using-pre-built-cellprofiler).
 
+If using RunDeepProfiler, check for alternative installation instructions in [RunDeepProfiler](RunDeepProfiler.md).
+
 ### Installing plugins without dependencies
 
 1. **Install CellProfiler.**  

setup.py

@@ -20,6 +20,11 @@
     "stardist"
 ]
 
+deepprofiler_deps = [
+    "numpy==1.23.0",
+    "inflect==6.0.0"
+]
+
 imagejscript_deps = [
     "pyimagej"
 ]
@@ -32,6 +37,7 @@
       "cellpose": cellpose_deps,
       "omnipose": omnipose_deps,
       "stardist": stardist_deps,
+      "deepprofiler":deepprofiler_deps,
       "imagejscript": imagejscript_deps,
     }
 )

tests/test_deepprofiler/config/profiling.json

@@ -0,0 +1,77 @@
+{
+    "dataset": {
+        "metadata": {
+            "label_field": "Metadata_Plate",
+            "control_value": "X"
+        },
+        "images": {
+            "channels": [
+                "DNA",
+                "RNA",
+                "ER",
+                "AGP",
+                "Mito"
+              ],
+            "file_format": "tif",
+            "bits": 16,
+            "width": 1224,
+            "height": 904
+        },
+	"locations":{
+	    "mode": "single_cells",
+        "area_coverage": 0.75,
+        "box_size": 128,
+        "mask_objects": false
+	}
+    },
+    "prepare": {
+        "illumination_correction": {
+            "down_scale_factor": 4,
+            "median_filter_size": 24
+        },
+        "compression": {
+            "implement": false,
+            "scaling_factor": 1.0
+        }
+    },
+    "train": {
+        "partition": {
+            "targets": [
+                "Metadata_Plate"
+            ],
+            "split_field": "Split",
+            "training": [0],
+            "validation": [1]
+        },
+        "model": {
+            "name": "efficientnet",
+            "crop_generator": "sampled_crop_generator",
+            "augmentations": true,
+            "metrics": ["accuracy", "top_k", "average_class_precision"],
+            "epochs": 3,
+	        "initialization":"ImageNet",
+            "params": {
+                "label_smoothing":0.0,
+                "learning_rate": 0.005,
+                "batch_size": 32,
+                "conv_blocks": 0
+            }
+        },
+        "sampling": {
+	    "factor": 1,
+            "workers": 1
+        },
+        "validation": {
+	    "frequency": 1,
+            "top_k": 5,
+            "batch_size": 32,
+            "frame": "val",
+            "sample_first_crops": true
+        }
+    },
+    "profile": {
+      "feature_layer": "block6a_activation",
+      "checkpoint": "Cell_Painting_CNN_v1.hdf5",
+      "batch_size": 32
+    }
+}

tests/test_deepprofiler/test_deepprofiler.cppipe

@@ -0,0 +1,128 @@
+CellProfiler Pipeline: http://www.cellprofiler.org
+Version:5
+DateRevision:425
+GitHash:
+ModuleCount:6
+HasImagePlaneDetails:False
+
+Images:[module_num:1|svn_version:'Unknown'|variable_revision_number:2|show_window:False|notes:['To begin creating your project, use the Images module to compile a list of files and/or folders that you want to analyze. You can also specify a set of rules to include only the desired files in your selected folders.']|batch_state:array([], dtype=uint8)|enabled:True|wants_pause:False]
+    :
+    Filter images?:Images only
+    Select the rule criteria:and (extension does isimage) (directory doesnot containregexp "[\\\\/]\\.")
+
+Metadata:[module_num:2|svn_version:'Unknown'|variable_revision_number:6|show_window:False|notes:['The Metadata module optionally allows you to extract information describing your images (i.e, metadata) which will be stored along with your measurements. This information can be contained in the file name and/or location, or in an external file.']|batch_state:array([], dtype=uint8)|enabled:True|wants_pause:False]
+    Extract metadata?:Yes
+    Metadata data type:Text
+    Metadata types:{}
+    Extraction method count:1
+    Metadata extraction method:Extract from file/folder names
+    Metadata source:File name
+    Regular expression to extract from file name:^(?P<Well>.*)_.*_.*_(?P<Site>.*)_(?P<Channel>.*)_001.tif
+    Regular expression to extract from folder name:(?P<Date>[0-9]{4}_[0-9]{2}_[0-9]{2})$
+    Extract metadata from:All images
+    Select the filtering criteria:and (file does contain "")
+    Metadata file location:Elsewhere...|
+    Match file and image metadata:[]
+    Use case insensitive matching?:No
+    Metadata file name:None
+    Does cached metadata exist?:No
+
+NamesAndTypes:[module_num:3|svn_version:'Unknown'|variable_revision_number:8|show_window:False|notes:['The NamesAndTypes module allows you to assign a meaningful name to each image by which other modules will refer to it.']|batch_state:array([], dtype=uint8)|enabled:True|wants_pause:False]
+    Assign a name to:Images matching rules
+    Select the image type:Grayscale image
+    Name to assign these images:DNA
+    Match metadata:[]
+    Image set matching method:Order
+    Set intensity range from:Image metadata
+    Assignments count:5
+    Single images count:0
+    Maximum intensity:255.0
+    Process as 3D?:No
+    Relative pixel spacing in X:1.0
+    Relative pixel spacing in Y:1.0
+    Relative pixel spacing in Z:1.0
+    Select the rule criteria:and (file does contain "DAPI")
+    Name to assign these images:DNA
+    Name to assign these objects:Cell
+    Select the image type:Grayscale image
+    Set intensity range from:Image metadata
+    Maximum intensity:255.0
+    Select the rule criteria:and (file does contain "GFP")
+    Name to assign these images:ER
+    Name to assign these objects:Nucleus
+    Select the image type:Grayscale image
+    Set intensity range from:Image metadata
+    Maximum intensity:255.0
+    Select the rule criteria:and (file does contain "Propidium")
+    Name to assign these images:AGP
+    Name to assign these objects:Cytoplasm
+    Select the image type:Grayscale image
+    Set intensity range from:Image metadata
+    Maximum intensity:255.0
+    Select the rule criteria:and (file does contain "CY5")
+    Name to assign these images:Mito
+    Name to assign these objects:Speckle
+    Select the image type:Grayscale image
+    Set intensity range from:Image metadata
+    Maximum intensity:255.0
+    Select the rule criteria:and (file does contain "CY5")
+    Name to assign these images:RNA
+    Name to assign these objects:Object1
+    Select the image type:Grayscale image
+    Set intensity range from:Image metadata
+    Maximum intensity:255.0
+
+Groups:[module_num:4|svn_version:'Unknown'|variable_revision_number:2|show_window:False|notes:['The Groups module optionally allows you to split your list of images into image subsets (groups) which will be processed independently of each other. Examples of groupings include screening batches, microtiter plates, time-lapse movies, etc.']|batch_state:array([], dtype=uint8)|enabled:True|wants_pause:False]
+    Do you want to group your images?:No
+    grouping metadata count:1
+    Metadata category:None
+
+IdentifyPrimaryObjects:[module_num:5|svn_version:'Unknown'|variable_revision_number:15|show_window:False|notes:[]|batch_state:array([], dtype=uint8)|enabled:True|wants_pause:False]
+    Select the input image:DNA
+    Name the primary objects to be identified:IdentifyPrimaryObjects
+    Typical diameter of objects, in pixel units (Min,Max):30,200
+    Discard objects outside the diameter range?:Yes
+    Discard objects touching the border of the image?:Yes
+    Method to distinguish clumped objects:Intensity
+    Method to draw dividing lines between clumped objects:Intensity
+    Size of smoothing filter:10
+    Suppress local maxima that are closer than this minimum allowed distance:7.0
+    Speed up by using lower-resolution image to find local maxima?:Yes
+    Fill holes in identified objects?:After both thresholding and declumping
+    Automatically calculate size of smoothing filter for declumping?:Yes
+    Automatically calculate minimum allowed distance between local maxima?:Yes
+    Handling of objects if excessive number of objects identified:Continue
+    Maximum number of objects:500
+    Use advanced settings?:Yes
+    Threshold setting version:12
+    Threshold strategy:Global
+    Thresholding method:Otsu
+    Threshold smoothing scale:1.3488
+    Threshold correction factor:1.0
+    Lower and upper bounds on threshold:0.0,1.0
+    Manual threshold:0.0
+    Select the measurement to threshold with:None
+    Two-class or three-class thresholding?:Three classes
+    Log transform before thresholding?:No
+    Assign pixels in the middle intensity class to the foreground or the background?:Background
+    Size of adaptive window:50
+    Lower outlier fraction:0.05
+    Upper outlier fraction:0.05
+    Averaging method:Mean
+    Variance method:Standard deviation
+    # of deviations:2.0
+    Thresholding method:Minimum Cross-Entropy
+
+RunDeepProfiler:[module_num:6|svn_version:'Unknown'|variable_revision_number:1|show_window:False|notes:[]|batch_state:array([], dtype=uint8)|enabled:True|wants_pause:False]
+    Select images to measure:AGP, DNA, ER, Mito, RNA
+    Input nuclei object name:IdentifyPrimaryObjects
+    Name this experiment.:FeaturesDeepProfiler
+    Output file location:Elsewhere...|C:\\Users\\fossa\\Desktop\\deepprofiler\\example\\run
+    DeepProfiler directory:Elsewhere...|C:\Users\fossa\DeepProfiler
+    Model directory:Elsewhere...|C:\Users\fossa\Desktop\deepprofiler\example
+    Model file:Cell_Painting_CNN_v1.hdf5
+    Select Metadata_Plate:Metadata_Frame
+    Select Metadata_Well:Metadata_Well
+    Select Metadata_Site:Metadata_Site
+    Config directory:Elsewhere...|C:\\Users\\fossa\\Desktop\\deepprofiler\\example
+    Model file:profiling.json