API

Import scBond:

import scBond

bond

bond.Bond()
bond.Bond.load_data(RNA_data, MET_data, ...)	Load data to Bond model.
bond.Bond.data_preprocessing([...])	Preprocessing for RNA data and methylation data in Bond.
bond.Bond.augmentation([enable_augmentation])	Data augmentation for Bond model (cell type only).
bond.Bond.construct_model(chrom_list[, ...])	Main model.
bond.Bond.train_model([R_encoder_lr, ...])	Training for model.
bond.Bond.test_model([model_path, ...])	Test for model.
bond.Bond.predict_single_rna([test_id_r, ...])	Predict methylation data from RNA data using existing test indices.
bond.Bond.predict_single_methylation([...])	Predict RNA data from methylation data using existing test indices.
bond.Bond.predict_single_modal([test_id, ...])	Single modal prediction for RNA or methylation data using existing test indices.

calculate_cluster

calculate_cluster.calculate_cluster_index(adata)

Evaluation of cluster index of prediction.

data_processing

data_processing.add_methylation_noise(MET_data)	Simulates measurement errors in methylation sequencing by randomly flipping a proportion of sites between 0 and 1.
data_processing.imputation_met(adata[, ...])	Impute missing values in methyaltion level matrices.
data_processing.MET_data_preprocessing(MET_data)	Preprocessing for methylation data, including filtering features, missing value imputation, optional noise addition, and normalization.
data_processing.RNA_data_preprocessing(RNA_data)	Preprocessing for RNA data, we choose normalization, log transformation and highly variable genes, using scanpy.

draw_cluster

draw_cluster.draw_tsne(data, title, color)	Visualization of prediction using tSNE Algorithm.
draw_cluster.draw_umap(data, title, color)	Visualization of prediction using UMAP Algorithm.

train_model

train_model.Model(RNA_data, MET_data, ...[, ...])
train_model.Model.predict_single_methylation([...])	Predict RNA data from methylation data using existing test indices.
train_model.Model.predict_single_rna([...])	Predict methylation data from RNA data using existing test indices.
train_model.Model.test(test_id_r, test_id_m)	Test for model.
train_model.Model.train(loss_weight, ...[, ...])	Training for model.