An engineer's review of over-the-counter CGMs for people without diabetes, covering sensor accuracy, normal reference ranges, how to get raw data out, and which metrics are worth computing.
A practical guide to wearing a continuous glucose monitor as a runner: sensor choice, raw data export, aligning glucose traces with .fit files, the metrics worth computing, and the failure modes that produce fake insights.
A step-by-step method for turning raw continuous glucose monitor exports into a reproducible report: data-sufficiency checks, core metrics, an ambulatory glucose profile, meal-anchored responses, and a written interpretation template you can hand to a clinician or an AI agent.
ABO and Rh status can be read directly from whole-genome sequencing data using four ABO variants and a copy-number check on RHD. This page gives the coordinates, the commands, and the failure modes that make a genotype call wrong.
A practical guide to pulling your lab results into a database, normalizing units and analyte identity, and deciding which changes between draws are signal rather than noise.
A working pipeline for bulk RNA-seq that stays in Python after quantification: salmon, pytximport, PyDESeq2, pycombat_seq, and gene set scoring, with the parameters and failure modes that matter.
A step-by-step pipeline from FASTQ files to a normalized expression matrix, differential expression across timepoints, and pathway-level summaries, with the flags and failure modes that matter for a single person's longitudinal data.
A working pipeline from FASTQ files to a ranked gene list and pathway enrichment, using salmon, tximport, DESeq2 and fgsea, with the quality control steps that matter for your own longitudinal blood samples.
A breakdown of bulk RNA-seq pricing by depth, library prep, and platform, plus what the quoted price leaves out and where the real cost of a usable experiment sits.
A concrete, runnable pipeline that takes FASTQ files from whole-genome and RNA sequencing through alignment, variant calling, and expression quantification, wrapped in Nextflow so it reruns cleanly.
The Yale calculator is Levine's PhenoAge: a nine-biomarker mortality model fit in NHANES III and rescaled into years. Here are the coefficients, the unit traps, the noise floor, and how we would track it over time.
Epigenetic age kits report a number derived from DNA methylation at a few hundred CpG sites. Here is how that number is computed, how much of it is noise, and what to measure instead if you want something you can track.
Biological age tests run from about $75 for a blood-chemistry clock to $300-$500 for a methylation array, with outliers above $1,000. Here is what each tier measures, how reliable it is, and how to compute the same clocks yourself from raw data.
A practical guide to choosing, timing, and interpreting blood and molecular markers for endurance and strength training, including how to tell a real change from analytical noise.
A practical guide to blood biomarker panels: which analytes carry information, how analytical and biological variation limit interpretation, and how to store and analyze your own results over time.
A single fasting glucose of 115 mg/dL is one noisy measurement of a tightly regulated system. Here is how to read it, what measurements resolve the ambiguity, and how to analyze the data yourself.
You can collect a sample at home, but you cannot get clinical-grade whole-genome data from a benchtop device. Here is what each option produces, what the files look like, and how to analyze them yourself.