Splicing code for microexons points to rare variants in autism cases without a known cause
Toronto, King's and Crick researchers built a splicing model that flags rare variants, mostly in autistic people without a genetic diagnosis. The same Nature paper scanned over 100 TB of public RNA-seq data and found more than 1,500 new microexons, GEN reported.
The Scientist · Science desk
85% of autistic individuals previously lacked a genetic link to autism, and variants predicted to affect microexons are mostly detected among them. Known variants give a genetic diagnosis in 15 percent of autistic individuals. More than 1,500 new microexons were found.
| Figure | Value | Note | Claim |
|---|---|---|---|
| Autistic people previously lacking a genetic link | 85% of autistic | 12 | |
| With a genetic diagnosis | 15% of autistic | 11 | |
| New microexons, more than | 1,500 microexons | 4 |
What happened
- The Nature abstract says the new microexons include many with neuronal subtype-specific splicing patterns and others that trigger nonsense-mediated mRNA decay.
- The team then built a machine-learning model to learn which regulatory features decide whether a microexon is spliced in or out of messenger RNA.
- They applied the model to whole-genome sequence data from thousands of people with autism and neurotypical sibling pairs, GEN reported.
- Blencowe's lab is also exploring new therapeutic approaches that target microexon dysregulation, according to GEN.
Why it matters
- What it limits constraint The flagged variants are predictions associated with autism status, so the roughly 15% diagnostic yield for known variants stays where it is until their splicing effects are tested.
- What's now possible capability Variants converging on genes not previously linked to autism give researchers a candidate list for follow-up experiments.
- What it sets up precedent The authors expect the same model to score newly found variants in schizophrenia and epilepsy, where microexons are believed to play a role.
Next story · Science · 5h ago
Baker lab's from-scratch zinc proteases cut their chosen target in 36% of 135 designsConfirmed2 publishers
Earlier microexon catalogues rest on a narrow kind of sample. According to the paper's introduction, previous studies analysed RNA-seq mostly from intact brain tissue, which can miss microexons spliced in specialized neuronal subtypes and those that trigger transcript instability [14]. The abstract reports both kinds among the new finds [13].
The counts differ by source. GEN, describing the analysis Parada ran with the MicroExonator tool he developed, puts the new microexons above 1,500, with nearly 300 appearing to be spliced specifically in brain [3][4]. The abstract says "hundreds" [13]. Taking GEN's figures at face value, brain-specific entries are at most about one in five of the new set [16].
"What surprised me was that the sequences the model identified as important closely matched those conserved throughout vertebrate evolution," Parada said [6]. In our view that agreement is the best evidence in the coverage that the model learned real regulatory sequence.
Blencowe said: "We were able to show that rare genetic variants generally have the largest predicted effects on the splicing of neuronal microexons" [8]. Those effects are scored from sequence, in whole-genome data from thousands of people with autism and their neurotypical siblings [7].
Blencowe put the clinical gap this way: "hundreds of genomic variants have been associated with autism, and these deliver a genetic diagnosis in 15 percent of autistic individuals," [11]. The variants predicted to affect microexons turn up mostly in the 85% of autistic people who previously lacked a genetic link [12]. But 85% is also the share of autistic people with no known genetic cause to begin with [10], so variants spread evenly across autistic individuals would land mostly there too [17].
The sibling pairs are the natural control [7], and the abstract says the variants are associated with autism status [9]. The thing this doesn't tell you is the size of that association. The sources do not report an effect size, how far above the 85% baseline the observed share sits, or any splicing measurement of the flagged variants in cells.
Blencowe named the next step in the lab's own terms: "In terms of future work, it will be important to understand how disrupted microexons affect interactions between proteins that are important for nervous system development and function," he said [18].
What to watch
- Splicing assays in neurons or animal models that measure whether the flagged variants actually change microexon inclusion.
- Application of the model to schizophrenia and epilepsy cohorts, which the authors say it can characterize.
- Blencowe's follow-up work on how disrupted microexons alter protein interactions in nervous system development.
Clarity's read
What the record supports and how the coverage leans. The claims behind it follow.
Reality
- Evidence72
- Adoption
- Not scored
- Hype gap+15
- Incentives
- Not scored
- Confidence70
Claim ledger
Ranked by verification strength, evidence, and original report placement.
- [1]
Scientists from the University of Toronto, the Sloan Kettering Institute, King's College London and the Francis Crick Institute published the findings in Nature.
- [2]
The scientists used a computational tool called MicroExonator to analyze over 100 terabytes of publicly available RNA sequencing data.
- [3]
MicroExonator was developed by Guillermo Parada, co-lead author and a postdoctoral fellow in Benjamin Blencowe's laboratory at King's and the Crick.
- [4]
Using the tool, Parada identified more than 1,500 new microexons, of which nearly 300 appeared to be specifically spliced in the brain.
- [5]
The scientists built a machine learning model designed to identify the key regulatory features governing whether a microexon is spliced in or out of messenger RNAs.
- [6]
Parada said: "What surprised me was that the sequences the model identified as important closely matched those conserved throughout vertebrate evolution."
- [7]
The scientists applied the model to whole-genome sequence data from thousands of people with autism and neurotypical sibling pairs.
- [8]
Blencowe said: "We were able to show that rare genetic variants generally have the largest predicted effects on the splicing of neuronal microexons."
- [9]
The variants predicted to alter microexon splicing are mostly detected in idiopathic individuals and are associated with autism status.
- [10]
High-confidence ASD-associated variants are detected in only around 15% of affected individuals; the remaining approximately 85% are idiopathic.
- [11]
Blencowe said: "hundreds of genomic variants have been associated with autism, and these deliver a genetic diagnosis in 15 percent of autistic individuals."
- [12]
The variants predicted to affect microexons are predominantly detected among the 85% of autistic individuals that previously lacked a genetic link to autism.
- [13]
The Nature abstract says transcriptome profiling detected hundreds of new microexons, including many with neuronal subtype-specific splicing patterns and others that activate nonsense-mediated mRNA decay.
- [14]
Previous studies focused primarily on analysing RNA-seq data from intact brain tissues, potentially missing microexons spliced in specialized neuronal subtypes or those that trigger transcript instability.
- [15]
The flagged variants converge on synaptic and other autism-linked genes, as well as genes not previously linked to autism.
- [16]
Nearly 300 brain-specific microexons out of more than 1,500 new ones is at most about 20 percent, roughly one in five.
- [17]
If flagged variants were spread evenly across autistic individuals, about 85% of them would fall in the idiopathic group, the same as that group's share of autistic individuals.
- [18]
Blencowe said: "In terms of future work, it will be important to understand how disrupted microexons affect interactions between proteins that are important for nervous system development and function."
- [19]
One of the research programs in Blencowe's lab is exploring new therapeutic approaches to target microexon dysregulation.
- [20]
The scientists believe the model can be used to characterize other newly discovered genetic variants and their effects in neurological conditions such as schizophrenia and epilepsy, where microexons are believed to play a role.
Sources
2 independent publishers whose own reporting we read for this story.
- genengnews.comNewly Identified Microexons Offer Clues to Autism Spectrum Disorder’s Genetic Roots
1 article · October 15, 2026
- nature.comA microexon regulatory code reveals autism-linked genetic variation
1 article · October 13, 2026
Topics and entities
Topics
- Alternative splicing
- Machine learning in genomics
Entities
- King's College London
- Francis Crick Institute
- Guillermo Parada
- University of Toronto
- Nature
- Benjamin Blencowe