MCP Tools Reference

mcp_pykingenie exposes the following tools to your AI assistant. These tools are intended only for surface-based binding data, such as Octet and Gator BLI experiments.

Data Import

Import tools accept absolute paths, or paths relative to the active MCP data directory shown by print_data_dir. Imported experiments are kept in memory for the current MCP session and can be checked with list_experiment_names.

Tool	Description
load_octet_example	Load the bundled Octet BLI example dataset.
import_octet_experiment	Import an Octet experiment from a folder of .frd sensor files and sample plate metadata.
import_gator_experiment	Import a Gator experiment from a folder or .zip file containing channel CSV files, Setting.ini, and ExperimentStep.ini.
import_kingenie_surface_csv	Import a KinGenie surface-simulation CSV file with trace and concentration columns.

Import Notes

import_octet_experiment

Pass a folder path. The folder should contain the Octet .frd files and sample plate files exported by the Octet software. If the path is not absolute, it is resolved inside the active MCP data directory.

import_gator_experiment

Pass either a folder path or a .zip archive. The input should include the assay channel CSV files plus Setting.ini and ExperimentStep.ini. Zip archives are extracted into the active MCP data directory, then loaded from the extracted folder.

import_kingenie_surface_csv

Pass a CSV file path. The file is expected to describe surface-based traces, with columns such as Time, Signal, Smax, and Analyte_concentration_micromolar_constant.

load_octet_example

Loads the packaged example as Example Experiment. This is useful for verifying that plotting, preprocessing, and fitting tools are working before importing your own data. Download the raw example files as octet_bli_example_data.zip.

File Utilities

Tool	Description
print_data_dir	Print the current output data directory path.
list_files_in_folder	List all files in a specified folder.

Experiment Inspection

Tool	Description
list_experiment_names	List all loaded experiment names.
list_experiment_properties	Get a property, such as sensor_names, across experiments.
list_experiment_attributes	List all attributes of a single experiment.
obtain_sample_info_table	Get analyte concentration and sensor metadata as JSON.

Data Processing

Tool	Description
align_association	Align the association phase of sensors.
align_dissociation	Align the dissociation phase of sensors.
subtract_reference	Subtract a reference sensor from other sensors.
subtract_experiment	Subtract one surface-based experiment from another sensor by sensor.
subtract_sensor_columns	Subtract paired sensor columns within one surface-based experiment.
align_and_subtract	Align and subtract a reference sensor in one step.

Plotting

Tool	Description
plot_sample_plate_info	Plot the sample plate layout for an experiment.
get_legends_table	Get the sensor legend DataFrame used by trace plots.
plot_traces_with_all_steps	Plot all BLI steps, including baseline, association, dissociation, and regeneration phases.
plot_kinetic_traces	Plot association and dissociation traces from fitting datasets.
plot_steady_state	Plot steady-state binding data.

Fitting

Tool	Description
initiate_fitting_datasets	Generate fitting datasets from a sample info JSON table.
run_steady_state_fitting	Run steady-state fitting as an alternative to kinetic fitting. Supports one_to_one and two_to_one.
run_kinetics_fitting	Run kinetic fitting as an alternative to standalone steady-state fitting. The tool handles PyKinGenie’s required starting values internally. one_to_one supports association_dissociation, association, and dissociation; one_to_one_mtl, one_to_one_if, and two_to_one support only association_dissociation.
get_kinetics_fitting_results	Retrieve fitting results, including Kd, k_off, Smax, and derived k_on.
create_export_df	Export raw or fitted association/dissociation trace points from generated fitting datasets as JSON records.