Thread by @patrick_griffN

Do you want a cheap (<$5) but highly accurate #aging biomarker that can be used in large experiments? Read our #preprint on TIME-Seq, a new method for inexpensive and scalable epigenetic age predictions. #clocks #longevity @davidasinclair biorxiv.org/content/10.110…

Epigenetic “clocks” based on DNA methylation are the most robust and widely used aging biomarker. However, conventional approaches (methylation chip or RRBS) are hundreds of $$ per sample and low throughput.

We wanted to routinely use epigenetic clocks in large studies for low cost. So, we developed Tagmentation-based Indexing for Methylation Sequencing (TIME-Seq) for scalable and inexpensive targeted methylation sequencing of DNA methylation biomarkers.

TIME-Seq leverages barcoded and bisulfite-compatible Tn5 transposomes to rapidly index sample DNA for a pooled library preparation that minimizes the cost of consumables ($0.65 / sample) and streamlines experiments with hundreds of samples.

TIME-Seq adaptors are also short and compatible with efficient hybridization enrichment of target epigenetic clock CpGs. This greatly reduces the cost of sequencing compared to RRBS, which has typically been used for sequencing-based clocks.

Using TIME-Seq, we first built an inexpensive (<$5) TIME-Seq epigenetic clock from 182 mouse blood DNA samples enriched for ribosomal DNA (rDNA).

Next, we built a clock from CpGs across the mouse genome that have high age correlation based on previous mouse clock studies. This clock costs slightly more (≈$10) but with testing could go much lower.

Around this time, our collaborators in @gladyshev_lab developed scAge, a probabilistic age-prediction algorithm to predict age from single cell methylation data. biorxiv.org/content/10.110…

We wondered if scAge could predict age in bulk samples with extremely shallow sequencing (e.g., 10K reads). To test, we shallow-sequenced 121 mouse blood TIME-Seq samples. The results were remarkable. Accurate prediction for <$2 / sample.

We validated the clocks and scAge-based approach in independent TIME-Seq library preps from new cohorts, as well as late-life dietary interventions. We also used TIME-Seq to make a human blood clock from hundreds of samples. Check out the preprint!

I hope TIME-Seq will allow researchers to use epigenetic clocks more regularly, in larger studies, and for way less $$. If you’d like to collaborate on a TIME-Seq clock, or use existing TIME-Seq clocks, reach out to me or @davidasinclair

@AlexandreTrapp and @gladyshev_lab have done further benchmarking of scAge + shallow sequencing by sampling existing RRBS data. Go check out their preprint too! biorxiv.org/content/10.110…

Thanks to @davidasinclair and all our fantastic collaborators:
@Alice_E_Kane
@AlexandreTrapp
Jien Li
@MaeveMcNamara3
Margarita Meer
@MikeMacArthur8
@sarahjayne189
@AmberMuellerPhD
Colleen Carmody
@_Daniel_Vera
@CsabaKerepesi
@nnhooten
@evansmkt
@gladyshev_lab