MOFA is a factor analysis model that provides a general framework for the integration of multi-omic data sets in an unsupervised fashion.
Intuitively, MOFA can be viewed as a versatile and statistically rigorous generalization of principal component analysis (PCA) to multi-omics data. Given several data matrices with measurements of multiple -omics data types on the same or on overlapping sets of samples, MOFA infers an interpretable low-dimensional representation in terms of a few latent factors. These learnt factors represent the driving sources of variation across data modalities, thus facilitating the identification of cellular states or disease subgroups.
In MOFA v2 (MOFA+) we added the following improvements:
-
Multi-group functionality: intuitively, this update breaks the assumption of independent samples and allows inference across multiple groups, where groups are predefined sets of samples (i.e. different conditions, batches, cohorts, etc.). Importantly, the model is not focused on capturing the differential changes between the groups (as for example when doing differential expression). The aim of the multi-group framework is to find out which sources of variability are shared between the different groups and which ones are exclusive to a single group.
-
GPU support: the training procedure can now be massively accelerated using GPUs. For this you have to install and configure the CuPy package.
For more details you can read our papers:
- MOFA v1: http://msb.embopress.org/cgi/doi/10.15252/msb.20178124
- MOFA+: https://www.biorxiv.org/content/10.1101/837104v1
The core of MOFA is implemented in Python. However, the whole procedure can be run with R and we provide the downstream analysis functions only in R.
Python dependencies can be installed using pip (from the Unix terminal)
pip install mofapy2MOFA2 R package can be installed using R:
remotes::install_github("bioFAM/MOFA2/MOFA2", build_opts = c("--no-resave-data --no-build-vignettes"))
# or use the devtools::install_github() equivalentYou can build an image using the provided Dockerfile:
docker build -t mofa2 .
You will then be able to use R or Python from the container.
docker run -ti --rm -v $DATA_DIRECTORY:/data mofa2 R
# ^
# |
# use `-v` to map a folder on your machine to a container directory
The command above will launch R with MOFA2 and its dependencies installed while mounting $DATA_DIRECTORY to the container.
You can also pull the pre-build image from dockerhub.
- Getting started: general overview and description of the method.
- Training a model in R
- Training a model in Python (jupyter notebook)
- Downstream analysis (in R)
- Downstream analysis in python: in preparation...
- Gene set enrichment analysis: demonstrates how to do gene set enrichment analysis.
- (authors' favourite) Analysis of chronic lymphocytic leukaemia cohort for personalised medicine: a bulk multi-omics data set. Figure 2 and 3 of the MOFA v1 paper.
- Analysis of a time course scRNA-seq data set using the multi-group framework: Figure 2 of the MOFA+ paper.
- Integration of single-cell multi-modal data (scNMT-seq): Figure 4 of the MOFA+ paper.
- Integration of single-cell multi-modal data (matching scRNA-seq and scATAC-seq)
- Analysis of CITE-seq data: in preparation...
- Analysis of microbiome data: in preparation...
We provide a Shiny-based web server to interactively explore MOFA models. Note that the web server only provides basic functionalities. For a comprehensive analysis please use the MOFA2 R package.
You can also download the latest version from the corresponding github repository
- Slack (recommended): we have a Slack group where we provide quick and personalised help, this is the link.
- e-mail: Ricard Argelaguet ([email protected])
(1.1) Can MOFA+ be applied to bulk data?
MOFA+ remains 100% applicable to bulk data.
(1.2) Does MOFA+ inherit previous features from MOFA v1?
Yes, pretty much everything: handling of missing data, non-gaussian likelihoods and sparsity in the weights.
(1.3) Do I need to provide multiple groups to use MOFA+?
No. Unless provided, MOFA+ assumes that you do not have multi-group structure in your data. In this case the model simplifies to MOFA v1 (but significantly faster).
(2.1) How does the multi-group inference work in MOFA+?
The aim of the multi-group framework is not to capture differential changes in mean levels between the groups (as for example when doing differential RNA expression). The goal is to compare the sources of variability that drive each group. If your aim is to find a factor that "separates" the groups, you DO NOT want to use the multi-group framework. In this setting, the features are centered per group before fitting the model.
(2.2) How do I define groups?
The selection of groups is hypothesis-driven, and typically motivated by the experimental design. There is no "right" or "wrong" definition of groups, but some definitions will be more useful than others. However, the user always needs to keep in mind that the aim of the multi-group framework is not to capture differential changes between the groups. The aim is to find out which sources of variability are shared between the different groups and which ones are exclusive to a single group. To achieve this, the group effect is regressed out from the data before fitting the model.
(2.3) How do I assess the quality/robustness of groups?
A quick approach to assess the validity of groups is to inspect the resulting variance explained plot. If the groups are too granular, the model will not recover significant amounts of variation. If the groups are not "interesting", this can result in a lack of "structure" in the variance explained plot (i.e. all factors being shared across all groups).
More computationally intensive approaches can be used to assess the robustness of groups, including cross-validation and downsampling or bootstrapping samples within groups.
(3.1) How do I normalise the data?
Proper normalisation is critical for the model to work. First, one needs to remove library size effects. For count-based data such as RNA-seq or ATAC-seq we recommend size factor normalisation + variance stabilisation. If this is not done correctly, the model will learn a very strong Factor 1 that will capture differences in the total expression per sample, and more subtle sources of variation will be downweighted.
(3.2) Should I do any filtering to the input data?
It is strongly recommended that you filter highly variable features (HVGs) per assay. When doing multi-group inference, you have to regress out the group effect before selecting HVGs.
(3.3) How many samples do I need?
Factor Analysis models are only useful with large sample sizes, at least more than ~25.
(3.4) Should I remove undesired sources of variability (i.e. batch effects) before fitting the model?
Yes. If you have clear technical factors, we strongly encourage to regress it out a priori using a simple linear model. The reason for this is that the model will "focus" on the huge variability driven by the technical factors, and smaller sources of variability could be missed. In prepare_mofa there is an argument called regress_covariates that you can use.
(3.5) My data sets have different dimensionalities, does this matter?
Yes. Bigger data modalities will tend to be overrepresented in the factors. It is good practice to filter uninformative features (based for example on a minimum variance threshold) in order to have the different views within the same order of magnitudes. If this is unavoidable, take into account that the model has the risk of missing (small) sources of variation in the small data set.
(3.6) Does MOFA handle missing values?
Yes. It simply ignores them from the likelihood, there is no hidden imputation step. Matrix factorisation models are known to be very robust to the presence of missing values!
(4.1) I get one of the following errors when running MOFA:
AttributeError: 'module' object has no attribute 'core.entry_point
Error in py_module_import(module, convert = convert) :
ModuleNotFoundError: No module named 'mofapy2'
First thing: restart R and try again. If the error still holds, this means that either you did not install the mofapy2 Python package (see instructions above), or you have multiple Python installations and R is not detecting the correct one where mofapy2 is installed. You need to find out the right Python interpreter (which usually will be the one you get when running which python in the terminal) and specify the following at the beginning of your R script:
library(reticulate)
use_python("YOUR_PYTHON_PATH", required=TRUE)
You can also use use_conda instead of use_python if you work with conda environments. For details read more about the reticulate package.
(4.2) I get the following error when installing the R package:
ERROR: dependencies 'XXX', 'YYY' are not available for package 'MOFA2'
You probably tried to install them using install.packages(). These packages should be installed from Bioconductor.
(4.3) I hate R, can I do MOFA only with Python?
You can use Python to train the model, see this notebook and this template script. However, we currently do not provide downstream analysis functions in Python (it is in our to-do list). For now we strongly recommend that you use our MOFA2 R package for this.
(4.4) Can I speed up the training procedure using CPU parallel processing? MOFA uses numpy for the mathematical operations. This library can be massively optimised by linking it to OpenBLAS or the Intel MKL libraries, which take advantage of multiple cores and multithreading.
You can check which libraries you have linked to numpy using
python -c "import numpy; print(numpy.show_config())"
Note that if you are using anaconda, numpy is automatically linked to MKL (see https://docs.anaconda.com/mkl-optimizations/).
The next step is to define the environmental variables. For MKL you need to set MKL_NUM_THREADS=N where N is the number of cores.
(4.5) How can I use GPUs to speed up training?
The Python core of MOFA can take advantage of NVIDIA GPUs to massively speed up training. For this you have to install and configure the CuPy package, which is an open-source matrix library accelerated with NVIDIA CUDA.
(5.1) How many factors should I learn?
The optimal number of factors depends on the aim of the analysis, the dimensions of the assays, the complexity of the data, there is no simple answer. In general, if the aim is to identify the major sources of biological variation one would typically consider the top 10 factors or so. In other tasks, such as imputation of missing values, even small sources of variation can be important and hence models should be trained with a large number of factors.
(5.2) Can I include known covariates in the model?
We extensively tested this functionality and it was not yielding good results. The reason is that covariates are usually discrete labels that do not reflect the underlying molecular biology. For example, if you introduce age as a covariate, but the actual age is different from the molecular age, the model will simply learn a new factor that corresponds to this latent molecular age, and it will drop the covariate from the model.
We recommend that you learn the factors in a completely unsupervised manner and then relate them to the biological covariates a posteriori (see vignettes). If your covariate of interest is an important driver of variability, do not worry, MOFA will find it!
(5.3) What data modalities can MOFA cope with?
- Continuous data: modelled using a gaussian likelihood
- Binary data: modelled using a bernoulli likelihood
- Count data: using a poisson likelihood
Importantly, the use of non-gaussian likelihoods require statistical approximations and are not as accurate as the gaussian likelihood. If your data can be safely transformed to match the gaussian likelihood assumptions, this is ALWAYS recommended. For example RNA-seq data is expected to be normalised and modelled with a gaussian distribution, do not input the counts directly.
(5.4) Do I need to do model selection?
As it occurs in most complex Bayesian models, the solution obtained depends on the parameter initialisation. In MOFA v1 we used random initialisation, which leads to (slightly) different solutions depending on the starting point. In MOFA v2 we initialise the factors using Principal Component Analysis on the concatenated data set, and the weights are initialised to zero. If using standard variational inference (not stochastic) this removes the randomness in the training algorithms, which guarantees a consistent solution.
(6.1) How can I do Gene Set Enrichment Analysis?
This is explained in the GSEA vignette
(6.2) How can I assess the robustness of factors? A procedure that can be applied to evaluate the robustness of factors is to downsample the number of samples and/or the number of features and inspect if the factors are consistently found. However, keep in mind that there could be cases where the full data set is required to detect small yet important sources of variation. Hence, lack of robustness under downsampling does not necessarily imply that a factor is not biologically meaningful.
@article{Argelaguet2018,
author = {Argelaguet, R. and Velten, B. and Arnol, D. and Dietrich, S. and Zenz, T. and Marioni, J. C. and Buettner, F. and Huber, W. and Stegle, O.},
title = {Multi-Omics Factor Analysis-a framework for unsupervised integration of multi-omics data sets},
journal = {Mol Syst Biol},
volume = {14},
number = {6},
pages = {e8124},
ISSN = {1744-4292 (Electronic) 1744-4292 (Linking)},
DOI = {10.15252/msb.20178124},
url = {https://www.ncbi.nlm.nih.gov/pubmed/29925568},
year = {2018},
type = {Journal Article}
}
@article {Argelaguet2019,
author = {Argelaguet, Ricard and Arnol, Damien and Bredikhin, Danila and Deloro, Yonatan and Velten, Britta and Marioni, John C and Stegle, Oliver},
title = {MOFA+: a probabilistic framework for comprehensive integration of structured single-cell data},
year = {2019},
doi = {10.1101/837104},
publisher = {Cold Spring Harbor Laboratory},
URL = {https://www.biorxiv.org/content/early/2019/11/09/837104},
journal = {bioRxiv}
}
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent