Bioinformatics Benchmarks¶

# Load entire dataset into memory - including expression matrix
adata = sc.read_h5ad("data.h5ad", backed="r")  # 7.8 MB for PBMC3K (metadata + expression)

# Filter cells using polars boolean indexing on metadata
filtered_cells = adata.obs.filter(pl.col("n_genes_by_counts") >= 500)

# Complex filtering with multiple conditions
high_quality = adata.obs.filter(
    (pl.col("n_genes_by_counts") >= 1000) &
    (pl.col("pct_counts_mt") <= 10)
)

# Cluster-based filtering
cluster_cells = adata.obs.filter(pl.col("leiden").is_in(["0", "1", "2"]))

# Load only metadata into memory
slaf = SLAFArray("data.slaf")  # Only loads obs/var metadata

# Direct filtering with SQL optimization
filtered_cells = slaf.filter_cells(n_genes_by_counts=">=500")

# Complex filtering with multiple conditions
high_quality = slaf.filter_cells(
    n_genes_by_counts=">=1000",
    pct_counts_mt="<=10"
)

# Cluster-based filtering
cluster_cells = slaf.filter_cells(leiden=["0", "1", "2"])

# Must load entire dataset
adata = sc.read_h5ad("data.h5ad", backed="r")

# Single-cell expression
cell_id = "AAACCTGAGAAACCAT-1"
cell_idx = adata.obs.index.get_loc(cell_id)
result = adata.X[cell_idx, :]

# Single-gene expression
gene_id = "MS4A1"
gene_idx = adata.var.index.get_loc(gene_id)
result = adata.X[:, gene_idx]

# Submatrix extraction
cell_start, cell_end = 0, 100
gene_start, gene_end = 0, 50
result = adata.X[cell_start:cell_end, gene_start:gene_end]

# No full dataset loading
slaf = SLAFArray("data.slaf")

# Single-cell expression
result = slaf.get_cell_expression("AAACCTGAGAAACCAT-1")

# Single-gene expression
result = slaf.get_gene_expression("MS4A1")

# Submatrix extraction
result = slaf.get_submatrix(
    cell_range=(0, 100),
    gene_range=(0, 50)
)

# Each step loads data into memory
adata = sc.read_h5ad("data.h5ad", backed="r")

# QC metrics calculation
sc.pp.calculate_qc_metrics(adata, inplace=True)

# Cell filtering with min_counts and min_genes
sc.pp.filter_cells(adata, min_counts=500, min_genes=200, inplace=True)

# Gene filtering
sc.pp.filter_genes(adata, min_counts=10, min_cells=5, inplace=True)

# Normalization
sc.pp.normalize_total(adata, target_sum=1e4, inplace=True)

# Log transformation
sc.pp.log1p(adata)

# Each operation processes the entire dataset
expression = adata.X[cell_ids, gene_ids]

# Build lazy computation graph
adata = LazyAnnData("data.slaf")  # LazyAnnData object

# QC metrics calculation (lazy)
pp.calculate_qc_metrics(adata, inplace=True)

# Cell filtering (lazy)
pp.filter_cells(adata, min_counts=500, min_genes=200, inplace=True)

# Gene filtering (lazy)
pp.filter_genes(adata, min_counts=10, min_cells=5, inplace=True)

# Normalization (lazy)
pp.normalize_total(adata, target_sum=1e4, inplace=True)

# Log transformation (lazy)
pp.log1p(adata)

# Only execute the computation on the slice of interest
expression = adata.X[cell_ids, gene_ids].compute()  # LazyExpressionMatrix.compute()