Linear Inference Models (iai_linear_inference)

The built-in model package iai_linear_inference trains a bundle of linear models for the target of interest against a specified list of predictors. It obtains the coefficients and variance estimates, and also calculates the p-values from the corresponding hypothesis tests. Linear inference is particularly useful for genome-wide association studies (GWAS), to identify genomic variants that are statistically associated with a risk for a disease or a particular trait.

This is a horizontal federated learning (HFL) model package.

Example Model Configuration

For binary targets, use the strategy LogitRegInference in the model_config.

For continuous targets, use LinearRegInference.

The data_config dictionary should include the following 3 fields (note that the columns in all the fields can be specified as either names/strings or indices/integers):

• target: the target column of interest;

• shared_predictors: predictor columns that should be included in all linear models (e.g., the confounding factors like age, gender in GWAS);

• chunked_predictors: predictor columns that should be included in the linear model one at a time (e.g., the gene expressions in GWAS)

// binary model example

model_config_logit = {
    "strategy": {"name": "LogitRegInference", "params": {}},
    "seed": 23,  # for reproducibility
}

// continuous model example
model_config_logit = {
    "strategy": {"name": "LinearRegInference", "params": {}},
    "seed": 23,  # for reproducibility
}

For more information, see Linear Inferences Sessions.