spatialtis.GCNG¶
- class spatialtis.GCNG(data, known_pairs=None, predict_pairs=None, train_partition=0.9, gpus=None, max_epochs=10, lr=0.0001, batch_size=32, random_seed=42, load_model=False, **kwargs)[source]¶
A pytorch reimplementation of GCNG
Use to identify directional gene-gene interactions. The trained model will be automatically save to anndata.
Note
To perform this analysis, you need pytorch, pytorch-geometry and pytorch-lightning installed. If you have GPU, make sure you install pytorch with GPU support, it would be way more faster than CPU.
- Parameters
data (anndata._core.anndata.AnnData) – AnnData object to perform analysis
known_pairs (Optional[pandas.core.frame.DataFrame]) – The input data for training, should be a dataframe with three columns, ligand, receptor, relationship; 0 means not interact, 1 means interact.
predict_pairs (Optional[List[Tuple]]) – The pairs that you interested
train_partition (float) – The ratio to split the dataset for training
gpus (Optional[int]) – Number of gpu to use, can be auto-detected
max_epochs (int) – Number of epoch
lr (float) – Learning rate
batch_size (int) – The batch size
random_seed (int) – The random seed
load_model (bool) – To load a pretrained model from anndata
**kwargs – Pass to
spatialtis.abc.AnalysisBase