Assuming you have a population with a condition and a population without this condition (control) then most studies use the classic comparison of averages using standard deviation.

This is easy to compute on paper or with most hand calculators. It also assume a gaussian distribution which is a very false assumption. With modern computers and software, additional methods are available — often better for detecting associations.

Some new contemporary methods

Doing queries for some of these methods illustrate that they are growing in use

For “random forest”

For “Mendelian randomization”

What is random forest?

Random forest is a powerful machine learning algorithm that creates an ensemble of multiple decision trees to produce more accurate and stable predictions. It can be used for both classification and regression tasks.

How Random Forest Works

1. Random sampling: The algorithm selects random samples from the training dataset.
2. Decision tree creation: For each sample, it constructs a decision tree.
3. Feature randomness: At each node of the tree, a random subset of features is considered for splitting, which helps reduce correlation between trees.
4. Prediction:
   • For classification: Each tree “votes” for a class, and the majority vote determines the final prediction.
   • For regression: The average of all tree outputs is taken as the final prediction.
5. Aggregation: The results from all trees are combined to produce the final output.

Random forest’s strength lies in its ability to reduce overfitting and handle complex datasets by leveraging the power of multiple uncorrelated decision trees

What is Mendelian randomization?

Mendelian randomization (MR) is a powerful epidemiological method that uses genetic variants as instrumental variables to investigate causal relationships between modifiable risk factors and health outcomes. This approach leverages the random assortment of genes during meiosis to mimic a randomized controlled trial, reducing the impact of confounding and reverse causation often present in observational studies.

How Mendelian Randomization Works?

1. Genetic variants selection: Researchers identify genetic variants strongly associated with the exposure of interest, typically from genome-wide association studies.
2. Instrumental variable analysis: These genetic variants serve as proxies or “instruments” for the exposure, allowing researchers to estimate its causal effect on the outcome.
3. Statistical analysis: The analysis often uses methods like two-stage least squares or summary data approaches, depending on whether individual-level or summary-level data is available.

By leveraging these principles and assumptions, Mendelian randomization provides a robust method for causal inference in epidemiology, complementing other forms of evidence in understanding disease etiology and potential interventions

Summary

In summary, while both methods aim to improve our understanding of relationships between variables, Mendelian randomization focuses on causal inference in epidemiology, whereas random forest is a versatile machine learning algorithm for prediction and classification tasks across various domains.

But wait!! There are still other ways that can be easily done with computers. One simple example using Pearson’s chi-squared test to test all viable percentile division of a dataset. For example, Neurocognitive: Brain Fog and we want to see what percentile threshold is best for the bacteria Alcaligenaceae.

Using 15%ile we have 1 out of 850 reporters with this level (and would expect 127). We can now test using 16%ile, 17%ile, etc to find the best threshold (highest statistically significance). to suggest brain fog is unlikely.

Similarly, at 85%ile we have 276 reporters (and would expect 127). We can now test using 86%ile, 84%ile, etc to find the best threshold (highest statistically significance). to suggest brain fog is likely.

With a full collection of different percentiles, we can generate some very interesting charts and using b-splines, functions.

I have heard from some microbiologists that they went into this disciple because they hated or was bad with mathematics… Mathematics has returned with a vengeance.