An MA plot puts fold change against mean expression; a volcano plot puts fold change against significance. Both are scatter plots of the same differential expression table, and each catches a different failure mode.
A log2 fold change of 1 means the measured quantity doubled between conditions. This page explains how that number is computed, why it is shrunk, and how to decide whether a doubling is real.
Log2 fold change is log2(treated / control). A log2FC of 4 is a 16-fold increase. This page covers the formula, the conversions, the code, and the failure modes that make a naive calculator wrong.
log2FC is the base-2 logarithm of a ratio between two expression estimates. This page explains the arithmetic, the shrinkage that DESeq2 and limma apply, and the failure modes that make a large log2FC meaningless.
A technical explanation of Olink's Proximity Extension Assay, what NPX values mean, how the panels and readouts differ, and how to handle the data files without misreading them.
A practical guide to Benjamini-Hochberg and Storey q-values on personal genomics, transcriptomics, and proteomics tables, including the diagnostics that tell you whether the correction is valid at all.
The genome is the fixed sequence, the transcriptome is which genes are being read right now, and the proteome is the working machinery. A practical comparison of what each layer measures, the file formats and tools involved, and why the three correlate less than most people expect.
A practical comparison of the three main molecular layers, including what each one measures, the file formats you get back, the failure modes, and how we would combine them in a personal dataset.
A working guide to turning a differential expression table from your own molecular data into ranked, deduplicated pathway results using fgsea, clusterProfiler, and GSVA, with the failure modes that make most enrichment output meaningless.
Genomics tells you what a cell could do, proteomics tells you what it is doing, and functional genomics is the question that connects them. A practical guide to the file formats, tools, and failure modes for someone working with their own multi-omic data.
A p-value is the error rate of one test under the null. An FDR of 0.05 is a property of the list you selected, meaning about 5% of the features you called significant are expected to be false. Here is how the two differ in practice, and how to compute them on your own omics data.
Genomics measures the DNA sequence you were born with. Proteomics measures the protein molecules present in a sample at one moment. This page explains the difference in terms of file formats, measurement noise, and what you can conclude from each.
A layer-by-layer guide to the multi-omics vendor market: instrument makers, service labs, assay platforms, and analysis software, judged by whether they hand you usable raw data.
A practical, opinionated toolchain for analyzing your own genome, transcriptome, proteome, and biomarker data: which tools to use at each layer, the flags that matter, and where the failure modes are.
There is no single best vendor. What matters is assay methodology, longitudinal consistency, and whether you can get structured data out. Here is how we compare the options and what a 100-biomarker panel really contains.
A precise account of the Benjamini-Hochberg step-up procedure: the arithmetic, the assumptions, the code, and the failure modes that matter when you run it on your own RNA-seq or proteomics data.
A curated list of blood, genomic, transcriptomic, proteomic, and glucose markers that carry real information about one person, with the variability numbers and file formats you need to work with them.
A practical survey of proteomics applications, the main measurement methods, and what you can do with your own plasma protein data once you have the files.
A practical pipeline for normalizing personal proteomics data, from raw intensity matrices through log transformation, method selection, batch correction, and longitudinal re-normalization when new draws arrive.
A practical comparison of mass spectrometry, affinity panels, and aptamer arrays as purchasable proteomics services, with the file formats, tools, and failure modes you should expect if you plan to analyze the data yourself.
A practical pipeline for turning vendor raw files from a plasma proteomics run into an FDR-controlled protein quantification matrix you can model, with the tools, flags, and failure modes named.
A step-by-step guide to processing personal plasma proteomics data, from raw mass spectrometry files or affinity panel exports through quantification, quality control, normalization, and longitudinal modeling in R.
A working list of blood, molecular, and continuous biomarkers worth tracking on yourself, with the biological-variation math that tells you when a change is real and when it is noise.
A working guide to blood biomarker analysis for technically fluent readers: what a single result can and cannot tell you, how to separate biological variation from signal, and how to structure your own longitudinal data.
A functional biomarker measures how well a system is working — a ratio, a rate, or a response to a challenge — rather than a single resting concentration. Here is how to build them from CGM, blood panels, proteomics, and RNA-seq, and where they break.
List prices for SomaScan 7K and 11K at core facilities run $800–900 per sample before shipping, plate minimums, and analysis. Here is where that money goes, how the assay works, and how to decide between SomaScan, Olink Reveal, and a targeted panel.
A Nature Medicine study staged 80 midlife women by hormone criteria and found 16 blood proteins that track the menopause transition better than age does. What the menopause proteomic score measures, how well it replicated, and what the dementia numbers do and do not support.
A plasma proteomics report measures the relative abundance of hundreds to thousands of circulating proteins. Here is what the numbers mean, how the three main platforms differ, and how to analyze your own data without fooling yourself.
A working guide to taking personal proteomics data from raw instrument files or NPX tables to a quantitative matrix, QC'd and normalized, and then to within-person statistics you can defend.
A step-by-step guide to taking a protein abundance matrix, building a defensible ranking statistic, and running over-representation and rank-based enrichment in R, including how to read a negative enrichment score.
A map of the multi-omics vendor market in four layers, what files each layer hands back, and how to evaluate a provider if you want data you can analyze yourself.
A technical guide to choosing and using a biomarker testing service: what analytes are worth measuring, what the analytical error looks like, how to store and version the results, and where a clinician has to be involved.
Dehydration concentrates plasma by a few percent, which is far too small to explain most ALT and AST elevations. Here is what usually causes them, how to draw a clean measurement, and how to read enzymes against your own genomic and proteomic baseline.
A working pipeline for taking Olink, SomaScan, or DIA-NN output into R: QC, missingness, normalization, and longitudinal within-person modeling of a single individual's plasma proteome.
A technical read of the MPMD/Marek Health blood panel: which assays are chosen well, which markers mislead in trained people, how to control pre-analytic noise, and where a large chemistry panel stops telling you anything new.