Dawnn is a method to detect differential abundance in a single-cell transcriptomic dataset.

Installation

Dawnn is currently only available from Github.

# Step 1: Install Dawnn package (may need to install `remotes` package first)
remotes::install_github("george-hall-ucl/dawnn")

# Step 2: Download Dawnn's model
# By default, model stored at ~/.dawnn/dawnn_nn_model.h5
dawnn::download_model()

# Step 3: Install Tensorflow Python package in Reticulate environment
reticulate::py_install("tensorflow")

Note: We are currently experiencing some installation issues on Apple-silicon (i.e. M1, M2, M3 chip) Macs, which we are trying to fix. See here.

Quick start

Assume that cells is a Seurat dataset with a PCA reduction, and a meta.data slot condition_name that contains the name of the condition to which each cell belongs (either Condition1 or Condition2). Dawnn requires at least 1,001 cells.

library(dawnn)

cells <- run_dawnn(cells, label_names = "condition_name", label_1 = "Condition1",
                   label_2 = "Condition2", reduced_dim = "pca")

After run_dawnn(), the object cells has additional meta.data slots:

Dawnn output	Description
cells$dawnn_scores	Output of Dawnn's model (estimated probability that a cell was drawn from sample with label_1)
cells$dawnn_lfc	Estimated log2-fold change in its neighbourhood.
cells$dawnn_p_vals	P-value associated with the hypothesis test that it is in a region of differential abundance.
cells$dawnn_da_verdict	Boolean output of Dawnn for whether it is in a region of differential abundance.

Dawnn's vignette explains these outputs in more detail.

Optional parameters

The above example only specifies the required parameters. Dawnn can be run in more complex scenarios by setting the following parameters:

cells <- run_dawnn(cells = cells, label_names = "condition_name",
                   label_1 = "Condition1", label_2 = "Condition2",
                   reduced_dim = "pca", n_dims = 20,
                   nn_model = "~/Documents/another_nn_model.h5,
                   recalculate_graph = FALSE, alpha = 0.025,
                   verbosity = 0, seed = 42)

These parameters are defined in the vignette.

Citation

Dawnn: single-cell differential abundance with neural networks. George T. Hall and Sergi Castellano (2023). Preprint on bioRxiv.

Contributions

Any contributions are warmly welcomed! Please feel free to submit an issue or pull request on this repository.

Licence

Copyright (C) 2023 University College London

This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program. If not, see http://www.gnu.org/licenses/.