pimms-learn 0.2.0

Imputing (MS-based prote-) omics data using self supervised deep learning models

PIMMS

PIMMS stands for Proteomics Imputation Modeling Mass Spectrometry and is a hommage to our dear British friends who are missing as part of the EU for far too long already (Pimms is also a British summer drink).

The pre-print is available on biorxiv.

PIMMS was called vaep during development.
Before entire refactoring has to been completed the imported package will be vaep.

We provide functionality as a python package, an excutable workflow and notebooks.

The models can be used with the scikit-learn interface in the spirit of other scikit-learn imputers. You can try this in colab.

Python package

For interactive use of the models provided in PIMMS, you can use our python package pimms-learn. The interface is similar to scikit-learn.

pip install pimms-learn

Then you can use the models on a pandas DataFrame with missing values. Try this in the tutorial on Colab:

Notebooks as scripts using papermill

If you want to run a model on your prepared data, you can run notebooks prefixed with 01_, i.e. project/01_*.ipynb after cloning the repository. Using jupytext also python percentage script versions are saved.

cd project # project folder as pwd
papermill 01_0_split_data.ipynb --help-notebook
papermill 01_1_train_vae.ipynb --help-notebook

Misstyped argument names won't throw an error when using papermill

Setup for PIMMS comparison workflow

The package is available as a standalone software on pypi. However, running the entire snakemake workflow in enabled using conda (or mamba) and pip to setup the environment. For a detailed description of setting up conda (or mamba), see instructions on setting up a virtual environment.

Download the repository

git clone https://github.com/RasmussenLab/pimms.git
cd pimms

Using conda (or mamba), install the dependencies and the package in editable mode

# from main folder of repository (containing environment.yml)
conda env create -n pimms -f environment.yml # slower
mamba env create -n pimms -f environment.yml # faster, less then 5mins

If on Mac M1, M2 or having otherwise issue using your accelerator (e.g. GPUs): Install the pytorch dependencies first, then the rest of the environment.

Install development dependencies

Check how to install pytorch for your system here.

• select the version compatible with your cuda version if you have an nvidia gpu

conda create -n vaep_manuel python=3.8 pip
conda activate vaep_manuel
conda update pytorch==1.13.1 torchvision==0.14.1 torchaudio==0.13.1 pytorch-cuda=11.7 -c pytorch -c nvidia # might be different
pip install . # pimms-learn
pip install papermill jupyterlab # use run notebook interactive or as a script
cd project
papermill 04_1_train_pimms_models.ipynb 04_1_train_pimms_models_test.ipynb # second notebook is output
python 04_1_train_pimms_models.ipynb # just execute the code
# jupyter lab # open 04_1_train_pimms_models.ipynb

Entire development installation

conda create -n pimms_dev -c pytorch -c nvidia -c fastai -c bioconda -c plotly -c conda-forge --file requirements.txt --file requirements_R.txt --file requirements_dev.txt
pip install -e . # other pip dependencies missing
snakemake --configfile config/single_dev_dataset/example/config.yaml -F -n

or if you want to update an existing environment

conda update  -c defaults -c conda-forge -c fastai -c bioconda -c plotly --file requirements.txt --file requirements_R.txt --file requirements_dev.txt

or using the environment.yml file (can fail on certain systems)

conda env create -f environment.yml

Troubleshooting

Trouble shoot your R installation by opening jupyter lab

# in projects folder
jupyter lab # open 01_1_train_NAGuideR.ipynb

## Run Demo

Change to the [`project` folder](./project) and see it's [README](project/README.md)

> Currently there are only notebooks and scripts under `project`, 
> but shared functionality will be added under `vaep` folder-package: This can 
> then be imported using `import vaep`. See [`vaep/README.md`](vaep/README.md)

You can subselect models by editing the config file:  [`config.yaml`](project/config/single_dev_dataset/proteinGroups_N50/config.yaml) file.

conda activate pimms # activate virtual environment cd project # go to project folder pwd # so be in ./pimms/project snakemake -c1 -p -n # dryrun demo workflow snakemake -c1 -p

The demo will run an example on a small data set of 50 HeLa samples (protein groups):
  1. it describes the data and does create the splits based on the [example data](project/data/dev_datasets/HeLa_6070/README.md)
     - see `01_0_split_data.ipynb`
  2. it runs the three semi-supervised models next to some default heuristic methods
     - see `01_1_train_collab.ipynb`, `01_1_train_dae.ipynb`, `01_1_train_vae.ipynb`
  3. it creates an comparison
     - see `01_2_performance_plots.ipynb`

The results are written to `./pimms/project/runs/example`, including `html` versions of the 
notebooks for inspection, having the following structure:

│ 01_0_split_data.html │ 01_0_split_data.ipynb │ 01_1_train_collab.html │ 01_1_train_collab.ipynb │ 01_1_train_dae.html │ 01_1_train_dae.ipynb │ 01_1_train_vae.html │ 01_1_train_vae.ipynb │ 01_2_performance_plots.html │ 01_2_performance_plots.ipynb │ data_config.yaml │ tree_folder.txt |---data |---figures |---metrics |---models |---preds

The predictions of the three semi-supervised models can be found under `./pimms/project/runs/example/preds`.
To combine them with the observed data you can run

```python
# ipython or python session
# be in ./pimms/project
folder_data = 'runs/example/data'
data = vaep.io.datasplits.DataSplits.from_folder(
    folder_data, file_format='pkl')
observed = pd.concat([data.train_X, data.val_y, data.test_y])
# load predictions for missing values of a certain model
model = 'vae'
fpath_pred = f'runs/example/preds/pred_real_na_{model}.csv '
pred = pd.read_csv(fpath_pred, index_col=[0, 1]).squeeze()
df_imputed = pd.concat([observed, pred]).unstack()
# assert no missing values for retained features
assert df_imputed.isna().sum().sum() == 0
df_imputed

Available imputation methods

Packages either are based on this repository, or were referenced by NAGuideR (Table S1). From the brief description in the table the exact procedure is not always clear.

Method	Package	source	status	name
CF	pimms	pip		Collaborative Filtering
DAE	pimms	pip		Denoising Autoencoder
VAE	pimms	pip		Variational Autoencoder
ZERO	-	-		replace NA with 0
MINIMUM	-	-		replace NA with global minimum
COLMEDIAN	e1071	CRAN		replace NA with column median
ROWMEDIAN	e1071	CRAN		replace NA with row median
KNN_IMPUTE	impute	BIOCONDUCTOR		k nearest neighbor imputation
SEQKNN	SeqKnn	tar file		Sequential k- nearest neighbor imputation starts with feature with least missing values and re-use imputed values for not yet imputed features
BPCA	pcaMethods	BIOCONDUCTOR		Bayesian PCA missing value imputation
SVDMETHOD	pcaMethods	BIOCONDUCTOR		replace NA initially with zero, use k most significant eigenvalues using Singular Value Decomposition for imputation until convergence
LLS	pcaMethods	BIOCONDUCTOR		Local least squares imputation of a feature based on k most correlated features
MLE	norm	CRAN		Maximum likelihood estimation
QRILC	imputeLCMD	CRAN		quantile regression imputation of left-censored data, i.e. by random draws from a truncated distribution which parameters were estimated by quantile regression
MINDET	imputeLCMD	CRAN		replace NA with q-quantile minimum in a sample
MINPROB	imputeLCMD	CRAN		replace NA by random draws from q-quantile minimum centered distribution
IRM	VIM	CRAN		iterativ robust model-based imputation (one feature at at time)
IMPSEQ	rrcovNA	CRAN		Sequential imputation of missing values by minimizing the determinant of the covariance matrix with imputed values
IMPSEQROB	rrcovNA	CRAN		Sequential imputation of missing values using robust estimators
MICE-NORM	mice	CRAN		Multivariate Imputation by Chained Equations (MICE) using Bayesian linear regression
MICE-CART	mice	CRAN		Multivariate Imputation by Chained Equations (MICE) using regression trees
TRKNN	-	script		truncation k-nearest neighbor imputation
RF	missForest	CRAN		Random Forest imputation (one feature at a time)
PI	-	-		Downshifted normal distribution (per sample)
GSIMP	-	script		QRILC initialization and iterative Gibbs sampling with generalized linear models (glmnet)
MSIMPUTE	msImpute	BIOCONDUCTOR		Missing at random algorithm using low rank approximation
MSIMPUTE_MNAR	msImpute	BIOCONDUCTOR		Missing not at random algorithm using low rank approximation
grr	DreamAI	-	Fails to install	Rigde regression
GMS	GMSimpute	tar file	Fails on Windows	Lasso model

Build status