SALRR: a pipeline that reads the brain's genetic book from cover to cover bioRxiv

Original: "SALRR: Scalable Analysis of Long-Read RNA-Seq Enables Comprehensive Transcriptome Profiling in Human Brain"
b42w:2608.90066 · the paper's page · 2026-08-29 · CC0 1.0 · 2 min
The automated SALRR platform turns long-read RNA sequencing of the brain from hands-on craft into a high-throughput production line.

In a print shop, every page of a book used to be set by hand. One mistake, and the whole print run is at risk. RNA sequencing of the brain looked much the same until recently: each sample demanded hours of manual work, and short reads, like torn scraps of pages, could not be pieced together into the full text of a gene. Now there is a production line that reads every page from start to finish.

The SALRR platform automates library preparation on a Hamilton robot, cutting hands-on time by 67% and raising throughput from 8 to 24 libraries per person per day. Its computational half, built in Snakemake, assembles long reads into full-length transcripts and classifies new isoforms. The result: across 13 libraries, an average of 68.8 million reads with a median length of 2,064 base pairs and a quality score of Q24.91.

The median poly(A) tail length of nuclear transcripts was 143 nucleotides, while that of mitochondrial transcripts was just 41, reflecting a known feature of mitochondrial mRNAs.

Long-read sequencing of postmortem brain used to be too labor-intensive and too sensitive to RNA degradation. Now it has been shown that SALRR performs reliably at RNA integrity numbers from 8.2 to 10 and yields reproducible isoform profiles. Expression correlation with short reads reached 0.742 at the gene level and 0.760 at the transcript level.

The platform detected 31,607 expressed isoforms from 10,075 genes, and 8,532 of them are new, which is 27%. Among the new isoforms, intron retention, novel combinations of known junctions and alternative 3' ends predominate. At the GBA1, CCNF, CHCHD10 and TREM2 loci, long reads resolved more isoforms than short reads, including two new GBA1 isoforms with an alternative 5' structure.

The platform is already ready for use on large postmortem brain cohorts, for example in the CARD Long-Read Sequencing Initiative, to hunt for isoforms linked to neurodegeneration.

A pipeline that reads every page of the genetic book in full, rather than in scraps, changes the rules of the game. Instead of manual typesetting, an automated line. Instead of scraps, the complete text. And now we can ask a question that was previously out of reach: how does isoform switching affect neuronal function in disease, when overall gene expression stays the same.

The median poly(A) tail length of nuclear transcripts was 143 nucleotides, while that of mitochondrial transcripts was just 41, reflecting a known feature of mitochondrial mRNAs.

In Greg Egan's novel Permutation City, copying a mind requires fully decoding the molecular states of neurons; SALRR brings reality closer to that level of detail by revealing isoform diversity that was previously invisible.

Take the whole work

Not published on arXiv. Here are its page, the PDF and the formalization: you can check it yourself, not only read our account.

Concepts
transcriptome transcriptomics long read sequencing brain scan PI-RADS classification