Breakthroughs

DeepMind’s AlphaFold2 (2021) and later AlphaFold3 (2023) solved the decades-old problem of accurately predicting 3D protein structures from amino acid sequences.

In 2021, DeepMind (yes, the people who made AI beat humans at Go) dropped AlphaFold2, and the scientific world basically lost its mind. This thing could predict a protein’s 3D structure with crazy-high accuracy — using only the amino acid sequence. No lab. No pipettes. Just AI and some mind-blowing math.

Then in 2023, they said: “Let’s go further.” And dropped AlphaFold3, which doesn’t just do proteins — it also predicts how proteins interact with DNA, RNA, ligands, and other molecules. It's like the difference between recognizing one LEGO piece and understanding the entire LEGO Death Star build at once.

A next-gen CRISPR tool that edits DNA with more precision and fewer off-target effects than classic CRISPR.

Prime editing is a next-generation genome editing technology first introduced in 2019 and significantly optimized through 2022–2023. It builds upon the CRISPR-Cas9 system by allowing targeted and programmable edits to DNA without inducing double-strand breaks or requiring donor DNA templates.

The system uses a modified Cas9 nickase fused to a reverse transcriptase enzyme, guided by a special "prime editing guide RNA" (pegRNA). This allows precise base substitutions, small insertions, and deletions at defined genomic loci.

A new reference genome built from 47 diverse individuals, replacing the old single "reference human."

The Human Pangenome is a next-gen reference genome, built from the complete DNA sequences of 47 genetically diverse individuals from around the world. Not just different continents — different ancestries, gene variants, and evolutionary histories.

Instead of one linear string of “this-is-human-DNA,” it’s a graph-based genome, mapping all the alternative paths DNA can take — kind of like Google Maps for genetics, showing not just the highway, but all the side roads, shortcuts, and scenic routes our genes can follow

Why it matters:

• The old genome missed up to 10% of human genetic variation — that's millions of base pairs missing from what we thought was “normal.”

• These missing pieces are not just trivia — they affect how we metabolize drugs, how diseases manifest, and even how genes are regulated.

So when researchers or doctors relied on the old reference to study diseases in non-European populations? Yeah... the data just wasn’t always accurate. It could lead to misdiagnosis, missed mutations, and inequities in research.

First CRISPR-Cas9 and CRISPR base editing treatments approved for sickle cell disease (Vertex/CRISPR Therapeutics) and β-thalassemia.

In 2023 and 2024, CRISPR finally got hired — not in a lab, not in a mouse, but in real human medicine. We’re talking full-on gene-editing treatments for people with sickle cell disease and β-thalassemia — two painful, inherited blood disorders that cause major suffering.

Here’s how it works: scientists take a patient’s own stem cells, use CRISPR-Cas9 or base editing (CRISPR’s more precise cousin) to fix or tweak the genes, and then put the edited cells back in. Boom — the cells start making healthy blood again.

And the craziest part? It works. Like, clinically-approved, regulatory-agency-said-yes kind of works.

Scientists created synthetic mouse and human-like embryo models from stem cells — no egg or sperm needed. 

In a series of breakthroughs between 2022 and 2024, scientists successfully created synthetic embryo-like structures—also called embryo models—from stem cells alone, no fertilized egg, no sperm, and no womb required. Using mouse and later human stem cells, researchers coaxed the cells to self-organize into structures that mimic real embryos, including the beginnings of organs, tissues, and even primitive brain-like features.

No genetic modification. No cloning. Just stem cells given the right signals, and nature took over.

Ethical alarm: These models aren't viable pregnancies — yet. But they’re close enough to raise real questions: What counts as life? Can we patent an embryo? Should we?

New biotech companies (like Shape Therapeutics, Korro Bio) are developing tools to edit RNA (not DNA) in living cells.

So we all know about DNA — it’s the holy grail, the master blueprint, the mothership of life’s instructions. And for years, gene editing has been all about going into that blueprint with molecular scissors like CRISPR, cutting, pasting, and rewriting DNA to fix inherited diseases.

But here’s the thing: what if you didn’t have to touch the DNA at all?
What if you could just... edit the message instead?

That’s exactly what RNA editing therapies aim to do.

RNA editing is a therapeutic strategy that modifies messenger RNA (mRNA) — the temporary molecules that carry instructions from DNA to make proteins. Instead of rewriting the genome itself, RNA editing corrects errors in the message, like autocorrect for cells.

Companies like Shape Therapeutics, Korro Bio, and others are developing platforms that use engineered enzymes (like ADARs) or programmable RNA editors to switch out single bases — for example, changing an A (adenosine) to an I (inosine), which cells read as G (guanosine).

In short: it’s precise, reversible, and doesn’t leave permanent edits in your DNA.