Skip to content

cloneSizebycellprofile

Function

cloneSizebycellprofile(
    self,
    gt_cna_file: str,
    tool_cna_files: List[str],
    tool_names: List[str],
    outfile: str = "clone_size_by_cell_profile.csv",
) -> pd.DataFrame

This function evaluates clone-size consistency at the cell profile level. It first derives clone sizes from the GT CNA matrix (cells sharing identical bin-level CNA profiles are grouped as one profile-clone), then compares each tool's inferred profile-clone sizes against GT.

The function writes:

  • a detailed per-cell comparison table
  • a mean summary grouped by GT cluster_size

Parameters

Name Type Description
gt_cna_file str Path to GT CNA profile CSV.
tool_cna_files List[str] List of tool CNA profile CSV files. Must align with tool_names in order.
tool_names List[str] Tool names used as output column prefixes ({Tool}_pred_size).
outfile str Detailed output filename. Default: "clone_size_by_cell_profile.csv".

Input File Format

gt_cna_file and each file in tool_cna_files are expected to be CNA matrices:

  • rows: genomic regions (first column as index when reading)
  • columns: cell IDs
  • values: CNA states (e.g., "1|1")

Example:

region,cell_001,cell_002,cell_003
chr1:1-100000,1|1,1|1,2|1
chr1:100001-200000,1|1,1|1,2|1

Implementation details from source:

  • missing values are filled with "1|1"
  • for tool matrices, all-empty columns are dropped before processing

Output

Two files are written to self.output_dir:

  1. os.path.join(self.output_dir, outfile)
  2. os.path.join(self.output_dir, f"mean_{outfile}")

1) Detailed Table (outfile)

Index: GT cell ID

Column Meaning
cluster_size GT profile-clone size for that cell
{Tool}_pred_size Predicted profile-clone size for the same cell ID from each tool (NaN if cell is missing in tool result)

2) Mean Table (mean_{outfile})

Grouped by GT cluster_size, numeric prediction columns are averaged:

Column Meaning
cluster_size GT profile-clone size
{Tool}_pred_size Mean predicted size for GT clones of that size

Return Type

The function signature annotates -> pd.DataFrame, but current implementation does not explicitly return a value (runtime return is None).

Example

from hcbench.gtbench.gtbench import GTBench

bench = GTBench(output_dir="out/gt_output")

bench.cloneSizebycellprofile(
    gt_cna_file="/path/to/gt/haplotype_combined.csv",
    tool_cna_files=[
        "/path/to/signals/haplotype_combined.csv",
        "/path/to/seacon/haplotype_combined.csv",
    ],
    tool_names=["signals", "seacon"],
    outfile="clone_size_by_cell_profile.csv",
)