Here we demonstrate how to integrate CITE-seq and scRNA-seq datasets with totalVI. The same principles here can be used to integrate CITE-seq datasets with different sets of measured proteins.
[1]:
import sys #if True, will install via pypi, else will install from source stable = True IN_COLAB = "google.colab" in sys.modules if IN_COLAB and stable: !pip install --quiet scvi-tools[tutorials] elif IN_COLAB and not stable: !pip install --quiet --upgrade jsonschema !pip install --quiet git+https://github.com/yoseflab/scvi-tools@master#egg=scvi-tools[tutorials]
|████████████████████████████████| 61kB 1.9MB/s ERROR: nbclient 0.5.0 has requirement jupyter-client>=6.1.5, but you'll have jupyter-client 5.3.5 which is incompatible. Installing build dependencies ... done Getting requirements to build wheel ... done Preparing wheel metadata ... done |████████████████████████████████| 163kB 2.8MB/s |████████████████████████████████| 122kB 11.7MB/s |████████████████████████████████| 112kB 11.4MB/s |████████████████████████████████| 8.7MB 5.3MB/s |████████████████████████████████| 3.2MB 72.1MB/s |████████████████████████████████| 2.4MB 68.6MB/s |████████████████████████████████| 7.7MB 15.0MB/s |████████████████████████████████| 51kB 8.4MB/s |████████████████████████████████| 51kB 7.7MB/s |████████████████████████████████| 112kB 64.1MB/s |████████████████████████████████| 51kB 7.7MB/s |████████████████████████████████| 61kB 8.7MB/s Building wheel for scvi-tools (PEP 517) ... done Building wheel for loompy (setup.py) ... done Building wheel for sinfo (setup.py) ... done Building wheel for numpy-groupies (setup.py) ... done
[2]:
import matplotlib.pyplot as plt import numpy as np import pandas as pd import plotnine as p9 import scanpy as sc import scvi sc.set_figure_params(figsize=(4, 4))
Here we focus on two CITE-seq datasets of peripheral blood mononuclear cells from 10x Genomics and used in the totalVI manuscript. We have already filtered these datasets for doublets and low-quality cells and genes.
The quality of totalVI’s protein imputation is somewhat reliant on how well the datasets mix in the latent space. In other words, it’s assumed here the datasets largely share the same cell subpopulations.
[3]:
adata = scvi.data.pbmcs_10x_cite_seq(run_setup_anndata=False)
INFO Downloading file at data/pbmc_10k_protein_v3.h5ad Downloading...: 24938it [00:00, 34828.85it/s] INFO Downloading file at data/pbmc_5k_protein_v3.h5ad Downloading...: 100%|██████████| 18295/18295.0 [00:00<00:00, 36564.20it/s]
/usr/local/lib/python3.6/dist-packages/anndata/_core/anndata.py:1094: FutureWarning: is_categorical is deprecated and will be removed in a future version. Use is_categorical_dtype instead if not is_categorical(df_full[k]): Observation names are not unique. To make them unique, call `.obs_names_make_unique`.
[4]:
# batch 0 corresponds to dataset_10k, batch 1 corresponds to dataset_5k batch = adata.obs.batch.values.ravel() adata.obs.batch
index AAACCCAAGATTGTGA-1 PBMC10k AAACCCACATCGGTTA-1 PBMC10k AAACCCAGTACCGCGT-1 PBMC10k AAACCCAGTATCGAAA-1 PBMC10k AAACCCAGTCGTCATA-1 PBMC10k ... TTTGGTTGTACGAGTG-1 PBMC5k TTTGTTGAGTTAACAG-1 PBMC5k TTTGTTGCAGCACAAG-1 PBMC5k TTTGTTGCAGTCTTCC-1 PBMC5k TTTGTTGCATTGCCGG-1 PBMC5k Name: batch, Length: 10849, dtype: object
Now we hold-out the proteins of the 5k dataset. To do so, we can replace all the values with 0s. We will store the original values to validate after training.
[5]:
held_out_proteins = adata.obsm["protein_expression"][batch == "PBMC5k"].copy() adata.obsm["protein_expression"].loc[batch == "PBMC5k"] = np.zeros_like(adata.obsm["protein_expression"][batch == "PBMC5k"])
[6]:
sc.pp.highly_variable_genes( adata, batch_key="batch", flavor="seurat_v3", n_top_genes=4000, subset=True )
/usr/local/lib/python3.6/dist-packages/anndata/_core/anndata.py:1094: FutureWarning: is_categorical is deprecated and will be removed in a future version. Use is_categorical_dtype instead if not is_categorical(df_full[k]): /usr/local/lib/python3.6/dist-packages/scanpy/preprocessing/_highly_variable_genes.py:144: FutureWarning: Slicing a positional slice with .loc is not supported, and will raise TypeError in a future version. Use .loc with labels or .iloc with positions instead. df.loc[: int(n_top_genes), 'highly_variable'] = True Observation names are not unique. To make them unique, call `.obs_names_make_unique`. Observation names are not unique. To make them unique, call `.obs_names_make_unique`.
Important
scvi-tools will automatically detect proteins as missing in a certain batch if the protein has 0 counts for each cell in the batch. In other words, to indicate a protein is missing in a certain batch, please set it to 0 for each cell.
[7]:
scvi.data.setup_anndata(adata, batch_key="batch", protein_expression_obsm_key="protein_expression")
INFO Using batches from adata.obs["batch"] INFO No label_key inputted, assuming all cells have same label INFO Using data from adata.X INFO Computing library size prior per batch
/usr/local/lib/python3.6/dist-packages/anndata/_core/anndata.py:1094: FutureWarning: is_categorical is deprecated and will be removed in a future version. Use is_categorical_dtype instead if not is_categorical(df_full[k]):
INFO Using protein expression from adata.obsm['protein_expression'] INFO Using protein names from columns of adata.obsm['protein_expression'] INFO Found batches with missing protein expression INFO Successfully registered anndata object containing 10849 cells, 4000 vars, 2 batches, 1 labels, and 14 proteins. Also registered 0 extra categorical covariates and 0 extra continuous covariates. INFO Please do not further modify adata until model is trained.
[8]:
vae = scvi.model.TOTALVI( adata, latent_distribution="normal", n_layers_decoder=2 )
[9]:
vae.train()
INFO Training for 400 epochs. INFO KL warmup for 8136.75 iterations Training...: 100%|██████████| 400/400 [07:46<00:00, 1.17s/it] INFO Training time: 444 s. / 400 epochs
[10]:
plt.plot(vae.trainer.history["elbo_test_set"], label="test") plt.title("Negative ELBO over training epochs") plt.ylim(1100, 1500) plt.legend()
<matplotlib.legend.Legend at 0x7fd315c1dc88>
Again, we rely on Scanpy.
[11]:
adata.obsm["X_totalVI"] = vae.get_latent_representation() adata.obsm["protein_fg_prob"] = vae.get_protein_foreground_probability(transform_batch="PBMC10k") rna, protein = vae.get_normalized_expression(transform_batch="PBMC10k", n_samples=25, return_mean=True)
Note
transform_batch is a power parameter. Setting this allows one to predict the expression of cells as if they came from the inputted batch. In this case, we’ve observed protein expression in batch “PBMC10k” (batch categories from original adata object), but we have no protein expression in batch “PBMC5k”. We’d like to take the cells of batch “PBMC5k” and make a counterfactual prediction: “What would the expression look like if my batch “PBMC5k” cells came from batch “PBMC10k”?”
transform_batch
[12]:
protein.iloc[:5, :5]
The following is for illustrative purposes. In the code blocks above, we have the denoised protein values for each cell. These values have the expected protein background component removed. However, to compare to the held out protein values, we must include both protein foreground and background. We recommend using the values above for downstream tasks.
[13]:
_, protein_means = vae.get_normalized_expression( n_samples=25, transform_batch="PBMC10k", include_protein_background=True, sample_protein_mixing=False, return_mean=True, )
[14]:
sc.pp.neighbors(adata, use_rep="X_totalVI") sc.tl.umap(adata, min_dist=0.4) sc.tl.leiden(adata, key_added="leiden_totalVI")
[15]:
perm_inds = np.random.permutation(len(adata)) sc.pl.umap( adata[perm_inds], color=["leiden_totalVI", "batch"], ncols=1, frameon=False, )
/usr/local/lib/python3.6/dist-packages/anndata/_core/anndata.py:1094: FutureWarning: is_categorical is deprecated and will be removed in a future version. Use is_categorical_dtype instead if not is_categorical(df_full[k]): /usr/local/lib/python3.6/dist-packages/anndata/_core/anndata.py:1192: FutureWarning: is_categorical is deprecated and will be removed in a future version. Use is_categorical_dtype instead if is_string_dtype(df[key]) and not is_categorical(df[key]) /usr/local/lib/python3.6/dist-packages/anndata/_core/anndata.py:1210: ImplicitModificationWarning: Initializing view as actual. "Initializing view as actual.", ImplicitModificationWarning Trying to set attribute `.obs` of view, copying. Observation names are not unique. To make them unique, call `.obs_names_make_unique`. Observation names are not unique. To make them unique, call `.obs_names_make_unique`. ... storing 'batch' as categorical
[16]:
batch = adata.obs.batch.values.ravel() combined_protein = np.concatenate([adata.obsm["protein_expression"].values[batch=="PBMC10k"], held_out_proteins], axis=0) # cleaner protein names parsed_protein_names = [p.split("_")[0] for p in adata.obsm["protein_expression"].columns] for i, p in enumerate(parsed_protein_names): adata.obs["{} imputed".format(p)] = protein_means.iloc[:, i] adata.obs["{} observed".format(p)] = combined_protein[:, i]
[17]:
viz_keys = [] for p in parsed_protein_names: viz_keys.append(p + " imputed") viz_keys.append(p + " observed") sc.pl.umap( adata[adata.obs.batch == "PBMC5k"], color=viz_keys, ncols=2, vmax="p99", frameon=False, add_outline=True, wspace=0.1 )
/usr/local/lib/python3.6/dist-packages/anndata/_core/anndata.py:1094: FutureWarning: is_categorical is deprecated and will be removed in a future version. Use is_categorical_dtype instead if not is_categorical(df_full[k]): /usr/local/lib/python3.6/dist-packages/anndata/_core/anndata.py:1192: FutureWarning: is_categorical is deprecated and will be removed in a future version. Use is_categorical_dtype instead if is_string_dtype(df[key]) and not is_categorical(df[key]) /usr/local/lib/python3.6/dist-packages/anndata/_core/anndata.py:1210: ImplicitModificationWarning: Initializing view as actual. "Initializing view as actual.", ImplicitModificationWarning Trying to set attribute `.obs` of view, copying. ... storing 'batch' as categorical
[18]:
from scipy.stats import pearsonr imputed_pros = protein_means[batch == "PBMC5k"] held_vs_denoised = pd.DataFrame() held_vs_denoised["Observed (log)"] = np.log1p(held_out_proteins.values.ravel()) held_vs_denoised["Imputed (log)"] = np.log1p(imputed_pros.to_numpy().ravel()) protein_names_corrs = [] for i, p in enumerate(parsed_protein_names): protein_names_corrs.append(parsed_protein_names[i] + ": Corr=" + str(np.round(pearsonr(held_out_proteins.values[:, i], imputed_pros.iloc[:, i])[0], 3))) held_vs_denoised["Protein"] = protein_names_corrs * len(held_out_proteins) held_vs_denoised.head()
We notice that CD15 has a really low correlation (imputation accuracy). Recall that imputation involves a counterfactual query – “what would the protein expression have been for these cells if they came from the PBMC10k dataset?” Thus, any technical issues with proteins in CD15 in PBMC10k will be reflected in the imputed values. It’s the case here that CD15 was not captured as well in the PBMC10k dataset compared to the PBMC5k dataset.
[19]:
p9.theme_set(p9.theme_classic) (p9.ggplot(held_vs_denoised, p9.aes("Observed (log)", "Imputed (log)")) + p9.geom_point(size=0.5) + p9.facet_wrap("~Protein", scales="free") + p9.theme(figure_size=(10, 10), panel_spacing=.35,) )
<ggplot: (8784187220572)>