Hi-C Compartments Calling (MCP-based)

Overview

This skill provides an automated workflow for compartments calling on .mcool, .cool or .hic Hi-C data.

Main steps include:

• Refer to the Inputs & Outputs section to verify required files and output structure.
• Always prompt user for genome assembly used.
• Always prompt user for resolution used to call compartments. ~50-250 kb is recommended. 100 kb is default.
• Locate the genome FASTA file from homer genome fasta file based on user input.
• Rename chromosomes in the .mcool or .cool file to satisfy the chromosome format with "chr".
• Generate chromosome-arm view files for compartment calling after changing the chromosome name.
• Perform PCA-based compartment analysis and extract the first principal component (PC1).
• Generate compartment interaction saddle plots and BigWig outputs for visualization.

When to Use This Skill

Use this skill when:

• You want to identify A/B compartments from Hi-C .mcool or .cool files.
• You need PC1 compartment scores and bigWig tracks for genome browser visualization.
• You want a reproducible, normalized, automated compartment-calling workflow.

Inputs & Outputs

Inputs

• File format: .mcool, .cool, or .hic (Hi-C data file) data.
• Genome assembly: Prompt the user for genome assembly used.
• Resolution: Prompt the user for resolution used to call compartments. The default resolution is 100 kb.

Outputs

${sample}_Compartments_calling/
    compartments/
      eigs.${resolution}.cis.vecs.tsv    # PC1 compartment scores  
      eigs.${resolution}.bw
      eigs.${resolution}.cis.lam.txt
      saddle.cis.${resolution}.digitized.tsv
      saddle.cis.${resolution}.saddledump.npz
    plots/         # PC1 track for genome browser  
      saddle.cis.${resolution}.pdf      # Saddle plot visualization 
    temp/
      expected.${resolution}.cis.tsv
      view_${genome}.tsv # Chromosome-arm view definition
      bins.${res}.tsv
      gc.${res}.tsv

Allowed Tools

When using this skill, you should restrict yourself to the following MCP tools from server cooler-tools, cooltools-tools, plot-hic-tools, project-init-tools, genome-locate-tools:

• mcpproject-init-toolsproject_init
• mcpgenome-locate-toolsgenomelocatefasta
• mcpHiCExplorer-toolshictomcool
• mcpcooler-toolslistmcoolresolutions
• mcpcooler-toolsharmonizechromnames
• mcpcooler-toolsmakeviewchromarms
• mcpcooler-toolsdumpbinsfor_gc
• mcpcooltools-toolsrungenomegc
• mcpcooltools-toolsrunexpectedcis
• mcpcooltools-toolsruneigscis
• mcpcooltools-toolsrun_saddle
• mcpplot-hic-toolsplotsaddlepdf

Do NOT fall back to:

• raw shell commands (cooler dump, cooltools eigs-cis, cooltools saddle, etc.)
• ad-hoc Python snippets (e.g. importing cooler, bioframe, matplotlib manually in the reply).

Decision Tree

Step 0 — Gather Required Information from the User

Before calling any tool, ask the user:

1. Sample name (sample): used as prefix and for the output directory ${sample}Compartmentscalling.

1. Genome assembly (genome): e.g. hg38, mm10, danRer11.

- Never guess or auto-detect.

1. Hi-C matrix path/URI (mcool_uri): e.g. .mcool file path or .hic file path.

- path/to/sample.mcool::/resolutions/100000 (.mcool file with resolution specified) - or .cool file path - or .hic file path

1. Resolution (resolution): default 100000 (100 kb).

- If user does not specify, use 100000 as default. - Must be the same as the resolution used for ${mcool_uri}

Step 1 — Initialize Project & Locate Genome FASTA

1. Make director for this project:

Call:

• mcpproject-init-toolsproject_init

with:

• sample: the user-provided sample name
• task: loop_calling

The tool will:

• Create ${sample}loopcalling directory.
• Return the full path of the ${sample}loopcalling directory, which will be used as ${proj_dir}.

1. If the user provides a .hic file, convert it to .mcool file using mcpHiCExplorer-toolshictomcool tool:

Call:

• mcpHiCExplorer-toolshictomcool

with:

• input_hic: the user-provided path (e.g. input.hic)
• sample: the user-provided sample name
• projdir: directory to save the view file. In this skill, it is the full path of the ${sample}loopcalling directory returned by mcpproject-init-toolsprojectinit.

The tool will:

• Convert the .hic file to .mcool file.
• Return the path of the .mcool file.

If the conversion is successful, update ${mcool_uri} to the path of the .mcool file.

1. Locate genome fasta file:

Call:

• mcpgenome-locate-toolsgenomelocatefasta

with:

• genome: the user-provided genome assembly

The tool will:

• Locate genome FASTA.
• Verify the FASTA exists.

Step 2: List Available Resolutions in the .mcool file & Modify the Chromosome Names if Necessary

1. Check the resolutions in mcool_uri:

Call:

• mcpcooler-toolslistmcoolresolutions

with:

• mcool_path: the user-provided path (e.g. input.mcool) without resolution specified.

The tool will:

• List all resolutions in the .mcool file.
• Return the resolutions as a list.

If the user defined or default ${resolution} is not found in the list, ask the user to specify the resolution again. Else, use ${resolution} for the following steps.

1. Check if the chromosome names in the .mcool file are started with "chr", and if not, modify them to start with "chr":

Call:

• mcpcooler-toolsharmonizechromnames

with:

• sample: the user-provided sample name
• projdir: directory to save the expected-cis and eigs-cis files. In this skill, it is the full path of the ${sample}Compartmentscalling directory returned by mcpproject-init-toolsprojectinit
• mcool_uri: cooler URI with resolution specified, e.g. input.mcool::/resolutions/${resolution}
• resolution: ${resolution} must be the same as the resolution used for ${mcool_uri} and must be an integer

The tool will:

• Check if the chromosome names in the .mcool file.
• If not, harmonize the chromosome names in the .mcool file.

Step 3 — Create Chromosome-Arm View File

Use bioframe to define chromosome arms based on centromeres:

Call:

• mcpcooler-toolsmakeviewchromarms

with:

• projdir: directory to save the expected-cis and eigs-cis files. In this skill, it is the full path of the ${sample}Compartmentscalling directory returned by mcpproject-init-toolsprojectinit
• mcool_uri: cooler URI with resolution specified, e.g. input.mcool::/resolutions/${resolution}
• resolution: ${resolution} must be the same as the resolution used for ${mcool_uri} and must be an integer
• genome: genome assembly

The tool will:

• Fetch chromsizes and centromeres via bioframe.
• Generate chromosomal arms and filter them to those present in the cooler.
• Return the path of the view file under ${proj_dir}/temp/ directory.

Step 4 — Compute GC Track for Bins

1. Dump bins for GC track:

Call:

• mcpcooler-toolsdumpbinsfor_gc

with:

• sample: the user-provided sample name
• projdir: directory to save the GC track file. In this skill, it is the full path of the ${sample}Compartmentscalling directory returned by mcpproject-init-toolsprojectinit
• mcool_uri: cooler URI with resolution specified, e.g. input.mcool::/resolutions/${resolution}
• resolution: ${resolution} must be the same as the resolution used for ${mcool_uri} and must be an integer

The tool will:

• Dump bins at the specified resolution from the cooler.
• Return the path of the bins file under ${proj_dir}/temp/ directory.

1. Compute GC track:

Call:

• mcpcooltools-toolsrungenomegc

with:

• sample: the user-provided sample name
• projdir: directory to save the GC track file. In this skill, it is the full path of the ${sample}Compartmentscalling directory returned by mcpproject-init-toolsprojectinit
• mcool_uri: cooler URI with resolution specified, e.g. input.mcool::/resolutions/${resolution}
• resolution: ${resolution} must be the same as the resolution used for ${mcool_uri} and must be an integer
• genome: genome assembly

The tool will:

• Compute GC content for each bin.
• Return the path of the GC track file under ${proj_dir}/temp/ directory.

Step 5 — Run Expected-cis and Eigs-cis (PCA Compartment Calling)

1. Calculate expected cis:

Call:

• mcpcooltools-toolsrunexpectedcis

with:

• sample: the user-provided sample name
• projdir: directory to save the expected-cis and eigs-cis files. In this skill, it is the full path of the ${sample}Compartmentscalling directory returned by mcpproject-init-toolsprojectinit
• mcool_uri: cooler URI with resolution specified, e.g. input.mcool::/resolutions/${resolution}
• resolution: ${resolution} must be the same as the resolution used for ${mcool_uri} and must be an integer
• viewpath: the path to the view file (e.g. ${projdir}/temp/view_${genome}.tsv)
• clrweightname: the name of the weight column (default: weight)
• ignore_diags: the number of diagonals to ignore based on resolution

The tool will:

• Generate expected cis file.
• Return the path of the expected cis file under ${proj_dir}/temp/ directory.

1. Calculate eigs cis:

Call:

• mcpcooltools-toolsruneigscis

with:

• sample: the user-provided sample name
• projdir: directory to save the expected-cis and eigs-cis files. In this skill, it is the full path of the ${sample}Compartmentscalling directory returned by mcpproject-init-toolsprojectinit
• mcool_uri: cooler URI with resolution specified, e.g. input.mcool::/resolutions/${resolution}
• resolution: ${resolution} must be the same as the resolution used for ${mcool_uri} and must be an integer
• viewpath: the view TSV from Step 3 (e.g. view${genome}.tsv)
• gc_tsv: GC track TSV from Step 4
• clrweightname: balancing column name (default "weight", but can be set based on clr.bins().columns if the user tells you the correct name)
• n_eigs: the number of principal components to compute (default 1)
• make_bigwig: whether to make bigwig file for PC1 track (default True)

This tool will:

• Run cooltools expected-cis to compute expected contact frequencies.
• Run cooltools eigs-cis to perform PCA and extract PC1.
• Return the path of the eigs-cis vecs file under ${proj_dir}/compartments/ directory.
• Return the path of the bigWig file under ${proj_dir}/compartments/ directory.

If the user reports an error about balancing weights:

• Ask the user which weight column should be used.
• Re-run expectedandeigs with the correct clrweightname.

Step 6 — Run Saddle Analysis

Call:

• mcpcooltools-toolsrun_saddle

with:

• sample: the user-provided sample name
• projdir: directory to save the saddle file. In this skill, it is the full path of the ${sample}Compartmentscalling directory returned by mcpproject-init-toolsprojectinit
• mcool_uri: cooler URI with resolution specified, e.g. input.mcool::/resolutions/${resolution}
• resolution: ${resolution} must be the same as the resolution used for ${mcool_uri} and must be an integer
• viewpath: the view TSV from Step 3 (e.g. view${genome}.tsv)
• eigsvecstsv: the eigs-cis vecs TSV from Step 5 (e.g. compartments/eigs.${resolution}.cis.vecs.tsv)
• expectedcistsv: the expected-cis TSV from Step 5 (e.g. temp/expected_cis.${resolution}.tsv)
• clrweightname: balancing column name (default "weight", but can be set based on clr.bins().columns if the user tells you the correct name)
• qrangelow and qrangehigh: default 0.02 and 0.98

The tool will:

• Run cooltools saddle.
• Generate saddle dump and related outputs, typically:
• Return the path of the saddle dump file under ${proj_dir}/compartments/ directory.
• Return the path of the other related outputs under ${proj_dir}/compartments/ directory.

Step 7 — Plot Saddle as PDF

Call:

• mcpplot-hic-toolsplotsaddlepdf

with:

• sample: the user-provided sample name
• projdir: directory to save the saddle file. In this skill, it is the full path of the ${sample}Compartmentscalling directory returned by mcpproject-init-toolsprojectinit
• resolution: ${resolution} must be the same as the resolution used for ${mcool_uri} and must be an integer
• chr_name: the user-provided chromosome name, e.g. chr1

This tool will:

• Load the corresponding .saddledump.npz file.
• Plot the saddle matrix with LogNorm(1e-1, 1e1) and RdBu_r colormap.
• Return the path of the compartment scores distribution PDF file under ${proj_dir}/plots/ directory.
• Return the path of the saddle plot PDF file under ${proj_dir}/plots/ directory.
• Return the path of the PC1 track PDF file under ${proj_dir}/plots/ directory.

If the saddledump file is missing, inform the user to run run_saddle first.

Best Practices

• Always confirm the genome and resolution explicitly with the user.
• Always use the defined MCP tools instead of ad-hoc code.
• If the user asks “how to run this manually”, you may conceptually describe the steps but still prefer to recommend using the MCP pipeline for reproducibility.
• If multiple resolutions are required, re-run the MCP tools with different resolution values and keep outputs in the same ${proj_dir} directory, using resolution in filenames for disambiguation.