antisplodge 0.1.3

AntiSplodge: A neural network-based spatial transcriptomics deconvolution pipeline

AntiSplodge

AntiSplodge, is a simple feed-forward neural network-based pipeline, designed to effective deconvolute spatial transcriptomics profiles, in an easy, fast, and, intuitive manner. It comes with all functions required to do a full deconvolution, from sampling synthetic spot profiles required to train the neural network, to the methods required to train the supplied network architecture. It is neatly packed into functions calls similar to that of traditional R-packages, where users are only exposed to fiddling with hyperparameters.

Installation

Using pip

You can install the package directy by running the following pip command:

pip install antisplodge

You can find the pip page at: https://pypi.org/project/AntiSplodge/

From GitHuB

You can install the package directly from GitHub by running the following command:

python -m pip install git+https://github.com/HealthML/AntiSplodge.git

Directly from source (this repository)

Clone the repository to a folder of your choice.

From a terminal this can be done by running:

git clone [e-mail]:HealthML/AntiSplodge.git

Subsequently, run the following pip command from your terminal (in the root of cloned directory):

pip install .

Usage

The full pipeline (see blow) assumes that you have a scRNA dataset (SC) and spatial transcriptomics dataset (ST) that both are formatted as .h5ad (AnnData data structures). Please see https://anndata.readthedocs.io/ for information about how to structure your data. Alternative you can check out the tutorial https://github.com/HealthML/AntiSplodge_Turorial for an example on how to do this.

Standard full pipeline

import AntiSplodge as AS
# SC should be the single-cell dataset formatted as .h5ad (AnnData)
Exp = AS.DeconvolutionExperiment(SC) 
Exp.setVerbosity(True)

# CELLTYPE_COLUMN should be replaced with actual column
Exp.setCellTypeColumn('CELLTYPE_COLUMN') 
# Use 80% as train data and split the rest into a 50/50 split validation and testing
Exp.splitTrainTestValidation(train=0.8, rest=0.5)

# Generate profiles, num_profiles = [#training, #validation, #testing]
# This will construct 10,000*10(CDs)=100,000, 5,000*10=50,000, 1,000*10=10,000 profiles  
# for train, validation and test (respectively)
Exp.generateTrainTestValidation(num_profiles=[10000,5000,1000], CD=[1,10])

# Load the profiles into data loaders
Exp.setupDataLoaders()

# Initialize Neural network-model and allocate it to the cuda_id specified
# Use 'cuda_id="cpu"' if you want to allocate it to a cpu
Exp.setupModel(cuda_id=6)
Exp.setupOptimizerAndCriterion(learning_rate = 0.001)

# Train the model using the profiles generated 
# The patience parameter determines how long it will run without fining a new better (lower) error 
# The weights found will be saved to 'NNDeconvolver.pt' and will be autoloaded once the training is complete 
stats = AS.train(Exp, save_file="NNDeconvolver.pt", patience=100)

# Check the testing accuracy
y_preds = AS.predict(Exp)

# Print test divergance as (Jensen Shannon divergance)
import numpy as np
from scipy.spatial import distance
jsds_ = []
for i in range(len(y_preds)):
    jsds_.append(distance.jensenshannon(Exp.Y_test_prop[i], y_preds[i]))
print("Mean {}".format(np.mean(jsds_)))

# Plot a boxplot of divergences for each test profile 
import seaborn as sns
import pandas as pd
pd.DataFrame({'jsds': jsds_}).to_csv("MouseBrainTestJSDS.csv")
sns.boxplot(y="JSD", data=pd.DataFrame({'JSD':jsds_}))

# Assuming we have a spatial transcriptomics dataset ST formatted in .h5ad (AnnData)
# create dataloader so that we can predict the profiles of each spot in our ST dataset
dataset_spots = AS.SingleCellDataset(torch.from_numpy(np.array(ST.X.toarray())).float(), torch.from_numpy(np.array([0]*ST.n_obs)).float())
spots_loader = DataLoader(dataset=dataset_spots,
                      batch_size=50, # batch_size doesn't matter 
)

spot_preds = AS.predict(Exp, spots_loader) # predict spots
# The results for each ST profile (spot) is now in spot_preds and can be used for further analysis 

Order of execution

1. Start an experiment: Exp = AS.DeconvolutionExperiment(SC) must be the first call.

2. Define datasets based on the SC dataset: Exp.splitTrainTestValidation(train=0.8, rest=0.5) must be called before Exp.generateTrainTestValidation(num_profiles=[10000,5000,1000], CD=[1,10]).

3. Setup model and optimizers: Exp.setupModel(cuda_id=6) must be called before Exp.setupOptimizerAndCriterion(learning_rate = 0.001). Each time setupModel is called, Exp.setupOptimizerAndCriterion must be called again, as optimizers and criterions are bound to the model, for use during training.

4. Train the model: stats = AS.train(Exp, save_file="NNDeconvolver.pt", patience=100).

5. Predict spots using the model: spot_preds = AS.predict(Exp, spots_loader).

The order of execution must be in the order listed above.

Useful snippets

Several ways of training

1. Standard training

The standard training procedure.

# Assuming an Exp is an DeconvolutionExperiment
AS.train(experiment=Exp, patience=25, save_file=None, auto_load_model_on_finish=True) # default parameters

2. Several warm restarts

Do 10 warm restarts with a low patience (n=5), this will autoload the model per train call. This will make the best model weights be loaded back onto the model and it will try again from these settings

best_error = None
# Do 10 warm restarts 
for i in range(10):
    AS.train(experiment=Exp, patience=5, best_loss=best_error)
    best_error = np.min(stats['validation_loss'])

3. Lowering learning rate

Start with a high learning rate and lower this by half for each warm restart.

lr = 0.01
all_stats = []
best_error = None

# do 5 warm restarts with decreasing learning rate
for i in range(5):
    print("Training with learning rate:", lr)
    Exp.setupOptimizerAndCriterion(learning_rate=lr)
    lr /= 10 # reduce learning rate by a factor of 10
    
    # For longer training, increase patience threshold
    stats = AS.train(Exp, save_file="NNDeconvolver.pt", patience=25, best_loss=best_error) 
    all_stats.extend(stats)
    
    best_error = np.min(stats['validation_loss']) # set best error as the target error to beat
    # the results in stats is the training errors during in each epoch (which might be needed for training plots)

4. Running on systems with reduced memory using smaller sets of training data

For users having trouble with the memory footprint of the profile generation, it is possible to generate smaller sets of training and validation profiles.

Exp.splitTrainTestValidation(train=0.8, rest=0.5) # define the dataset splits
Exp.setupModel(cuda_id=6) # the model can be built beforehand
best_error = None

# do 100 warm restarts with smaller chunks of training data
for i in range(100):
    Exp.generateTrainTestValidation(num_profiles=[5000,1000,1], CD=[1,10])
    Exp.setupDataLoaders()
    AS.train(experiment=Exp, save_file="CurrentDeconvolver.pt", patience=10, best_loss=best_error)
    
    best_error = np.min(stats['validation_loss']) 

# Remember to generate test profiles after training is complete 
Exp.generateTrainTestValidation(num_profiles=[1,1,1000], CD=[1,10])

# Continue as usual

Tutorial

Check out the tutorial located at: https://github.com/HealthML/AntiSplodge_Turorial. This will give you a full tour from preprocessing to deconvoluting by predicting cell type proportions of the spatial transcriptomics spots.

Dependencies

The list of major dependencies are:

• numpy>=1.17.2 (https://numpy.org/)
• pandas>=0.25.3 (https://pandas.pydata.org/)
• scikit-learn>=0.22.1 (https://scikit-learn.org/)
• torch>=1.9.0 (https://pytorch.org/)

Documentation

The documentation is available at: https://antisplodge.readthedocs.io/.

References

Coming soon.

License

The source code for AntiSplodge is licensed under the MIT License. See the LICENSE file for details.

Known issues

• AntiSplodge is prone to be affected by bad initiations. Oftentimes, this can be resolved by simply restarting the Experiment (or re-initializing the model). This seems to be more frequent when solving problems with many classes (large number of cell types). If verose is set to true, you should see output warnings during training with (!!NaNs vectors produced!!, these are not a problem if they only persist for a single iteration and is gone in the next).