Molecular biology studies DNA, RNA and proteins and the flow of information between them. Here are the basics explained once, properly, with the techniques and why anyone analysing sequencing data needs them.
Category: Bioinformatics
NGS in Cancer: How Tumour Sequencing Works and What It Finds
Cancer sequencing compares a tumour against the patient’s own healthy DNA. Here is what that finds, why tumour samples are harder to analyse and how the pipeline runs from sample to report.
Whole-Genome vs Whole-Exome Sequencing for Cancer: How to Choose
Why tumour purity and subclonality make depth decisive, what only WGS can detect, where panels win, and the matched-normal decision that matters more than any of it.
ATAC-Seq Quality Control: TSS Enrichment, Fragment Size and FRiP Explained
The three metrics that decide whether an ATAC-Seq library is usable, how to read the fragment size distribution, and how to diagnose exactly which step of the protocol failed.
Cell Type Annotation in Single-Cell RNA-Seq: How to Get It Right
Choosing clustering resolution, triangulating annotation evidence, spotting doublets and ambient RNA, and the replicate error that invalidates most single-cell differential expression.
Computational Enzyme Design and Engineering: What Works, and What Still Needs the Bench
Stabilising an enzyme, changing its specificity, and designing catalysis from scratch are three different problems with very different odds. What computation contributes to each.
Large Language Models in Bioinformatics: What They Are Actually Good For
Protein language models and text LLMs do very different jobs. Where embeddings genuinely outperform, where generation fails, and the evaluation error that inflates most published results.
De Novo Protein Design with RFdiffusion and ProteinMPNN: What Actually Works
How modern de novo protein design works, the real success rates, where it beats antibodies, and the filters that decide whether a campaign succeeds.
16S rRNA vs Shotgun Metagenomics: How to Choose the Right Approach
What each method actually measures, the biases built into both, what they cost, and a straightforward rule for deciding which your project needs.
Reverse Vaccinology: How Computational Vaccine Design Works, Step by Step
From pathogen genome to a codon-optimised multi-epitope construct — the full immunoinformatics pipeline, and an honest account of which steps you can trust.
