02 / RESEARCH TODAY
New today.
On bioRxiv.
Selected preprints first posted on November 18, 2026, in areas close to Algenom’s focus: genomics, computational biology, synthetic biology and plant microbiology.
GENOMICS · PREPRINT
Jumping genes leave a traceable record of family history among human brain cells
What the authors report. Demirci and colleagues, with a corresponding author at the Karolinska Institute, introduce LUSTRE, a single-cell assay that maps new insertions of the L1 retrotransposon without amplifying the whole genome. Profiling about 17,600 neurons from the dentate gyrus, cortex and cerebellum of adult human brains, they estimate, after validation, 0.1–0.3 new somatic L1 insertions per neuron. Shared insertions mark related cell lineages, and the data point to two separate waves of L1 activity during development. Tracing sequence carried along with the insertions, they find that a small number of source L1 copies account for most events.
Why it matters. L1 is the only element in the human genome still able to copy itself autonomously, and how often it does so in the brain has been hard to measure. A scalable assay turns these insertions into natural barcodes for reconstructing how cells in a human tissue are related.
The open question. The rate is described as a conservative, validation-adjusted estimate from post-mortem adult tissue, and the abstract does not address whether the insertions affect how neurons work. Lineage relationships are inferred rather than directly observed.
COMPUTATIONAL BIOLOGY · PREPRINT
An AI model reads isotope-labelling data to estimate how fast cells run their metabolic reactions
What the authors report. Weilandt and colleagues, with a corresponding author at Princeton University, present Flux Transformer, a model that uses reaction-specific attention to combine isotope-labelling patterns with measured reaction rates and output mass-balanced, steady-state metabolic fluxes. Trained on simulated data with added noise, it recovered the underlying fluxes more effectively than traditional metabolic flux analysis. On experimental data from wild-type and mutant bacteria, yeast, plants and mammalian T cells, it reproduced fluxes that experts had curated by hand.
Why it matters. Flux analysis is central to metabolic engineering but is slow to compute and often needs expert tuning, which keeps it out of reach for many labs. A model that does the inversion quickly could make the method routine.
The open question. The model learns from simulations, so it may struggle with networks or conditions unlike its training data. Agreement with expert-curated results shows consistency with current practice rather than independent proof that the fluxes are correct.
SYNTHETIC BIOLOGY · PREPRINT
A plug-in plasmid kit to turn genes up, down and back again in a photosynthetic bacterium
What the authors report. Guevara and colleagues, at the Complutense University of Madrid, characterised a set of inducible promoters on standardised SEVA plasmids in the cyanobacterium Synechocystis sp. PCC 6803, using fluorescent reporters to compare their range, strength and leakiness. They also built a single-plasmid CRISPR interference (CRISPRi) vector in which a catalytically dead Cas9 sits under a tightly controlled promoter combining a theophylline riboswitch with rhamnose induction, alongside its guide RNA. As a proof of concept, the system reversibly knocked down a chromosomal yellow fluorescent reporter.
Why it matters. Cyanobacteria fix carbon dioxide and need few nutrients, which makes them attractive production hosts. Compact, well-characterised parts for switching genes on and off are a basic requirement for engineering them reliably.
The open question. The knockdown was shown on a reporter gene, not on a native metabolic gene or a production pathway. How the parts behave at industrial scale or in other cyanobacterial strains is not tested here.
PLANT MICROBIOLOGY · PREPRINT
A root fungus that helps plants grow keeps working in salty conditions
What the authors report. Gonzalez-Sayer and colleagues, with a corresponding author at the Centre for Plant Biotechnology and Genomics (UPM-CSIC) in Madrid, assembled a near-chromosome-level, 29.39 Mb genome for Penicillium melinii isolate 2, a root endophyte of Arabidopsis, along with what they describe as the species’ first complete mitochondrial genome. Compared with two other growth-promoting Penicillium species, it encodes fewer carbohydrate-active enzymes, biosynthetic gene clusters and candidate effectors. In lab tests it showed 7 of 11 growth-promoting traits, including phosphate and zinc solubilisation, kept growing at high salt, and increased Arabidopsis shoot and root growth on medium containing 100 mM sodium chloride.
Why it matters. Soil salinity limits crop yields in many regions. Microbes that support plant growth under salt stress are candidates for biological inoculants, and a high-quality genome gives a starting point for finding the genes responsible.
The open question. The plant results come from Arabidopsis on lab medium, not crops in saline fields. The candidate genes were identified by genome mining and still need functional validation, as the authors note.