The age of prenatal genetic screening is here (let’s call it that!)

In the spring of 2010, I went to the studios of KQED in San Francisco to record an interview with a radio show on the BBC about PGD. Preimplantation genetic diagnosis. I haven’t thought much about the issue in the near ten years since then. Which in a personal sense certainly reflects my luck and circumstance.

But I’m thinking about the issue after reading this story from Emily Mullin, We’re Already Designing Babies: Expanded genetic testing of embryos represents a new era of family planning. But how far should the technology go?:

JJill Pinarowicz’s life has been shaped by a mutation in her mother’s DNA. The genetic error gave her two brothers a rare disease called Wiskott-Aldrich syndrome….

Both of Pinarowicz’s brothers passed away from complications of the disease. One died as a toddler, before she was born, and her other brother died at age 18, when Pinarowicz was a teenager.

Pinarowicz thought it would be too risky to have her own children….

The technique is called preimplantation genetic testing (PGT). By using PGT together with in-vitro fertilization, Pinarowicz and her husband had a healthy son in May 2017.

An incredible “feel-good” outcome so far. And not surprising. I have become more conservative about technology since I first started writing on the internet in the early 2000s, but I will never oppose these sorts of genetic technologies that allow couples whose offspring are at high risk of developing serious debilitating conditions to avoid these scenarios. But the magnitude of how common this now took me aback:

The U.S. Centers for Disease Control and Prevention (CDC) reported in January that PGT was used in 22 percent of IVF cases in 2016, up from just 5 percent in the previous year.

Since the last statistic Mullin could find was from 2016, it’s almost certain that the proportion is greater than 22 percent today. The numbers for 2018 seem difficult to find, but it seems likely that ~75,000 live-births per year in the USA are now due to IVF. Worldwide there are in the range of 10 million humans alive today due to IVF.

How relevant IVF is to fertility varies by social and demographic variables. I know a fair number of people who have done IVF. The average age of a mother at her first birth is 32 in San Francisco and 31 in Manhattan. As many of you probably know many options relating to fertility and genetic testing come “online” for American insurance companies at age 35.

When you transform blue-sky exotic basic science into mass technology they become far less controversial. One of the major themes of Carl Zimmer’s new book, She Has Her Mother’s Laugh, was the vocal and mainstream nature of 20th-century eugenics. A major criticism of Robert Plomin’s Blueprint is that it was resurrecting genetic determinism. Let me quote Mullin:

In Iceland, for instance, the widespread availability of prenatal genetic testing has meant that nearly 100 percent of women choose to abort a fetus with Down syndrome, which has led to a near eradication of babies being born with the condition.

What is in a word? Something in the future is worrisome. Something that professional dual-income-no-kids couples do in their attempt to attain the classic bourgeois lifestyle is not so worthy of comment? Outside of the pro-life movement the discussion of the ubiquity of screening for Down syndrome seems rather muted, even though it is widespread. While we may furrow our brows over decisions made based on polygenic risk scores, the reality is that the age of Mendelian screening is here. It is not speculative science, but applied medicine.

Call it what you want to call it.

Release the UK Biobank! (the prediction of height edition)


There’s so much science coming out of the UK Biobank it’s not even funny. It’s like getting the palantír or something.

Anyway, a preprint, submitted for your approval. A vision of things to come? Accurate Genomic Prediction Of Human Height:

We construct genomic predictors for heritable and extremely complex human quantitative traits (height, heel bone density, and educational attainment) using modern methods in high dimensional statistics (i.e., machine learning). Replication tests show that these predictors capture, respectively, ~40, 20, and 9 percent of total variance for the three traits. For example, predicted heights correlate ~0.65 with actual height; actual heights of most individuals in validation samples are within a few cm of the prediction. The variance captured for height is comparable to the estimated SNP heritability from GCTA (GREML) analysis, and seems to be close to its asymptotic value (i.e., as sample size goes to infinity), suggesting that we have captured most of the heritability for the SNPs used. Thus, our results resolve the common SNP portion of the “missing heritability” problem – i.e., the gap between prediction R-squared and SNP heritability. The ~20k activated SNPs in our height predictor reveal the genetic architecture of human height, at least for common SNPs. Our primary dataset is the UK Biobank cohort, comprised of almost 500k individual genotypes with multiple phenotypes. We also use other datasets and SNPs found in earlier GWAS for out-of-sample validation of our results.

A scatter-plot is worth a thousand derivations.

You know what better than 500,000 samples? One billion samples! A nerd can dream….