Marc Robinson-Rechavi
LightNews LightNews
banner
marcrr.bsky.social
Marc Robinson-Rechavi
@marcrr.bsky.social
1.6K followers 1.2K following 340 posts
He/Him. Evolution and Bioinformatics. Posts mine alone, in English (mostly) & French. Chair of @dee-unil.bsky.social, Prof at @fbm-unil.bsky.social‬, group leader at @sib.swiss. PI of @bgee.org
🦣 Main account: https://ecoevo.social/@marcrr
Posts Replies Media Videos
marcrr.bsky.social
The 2025 state of the climate report: a planet on the brink
doi.org/10.1093/bios...
Approximate global average temperature from –9340 BCE to 2020 CE, showing a gradual increase then decrease until fossil fuel use accelerates ≈1850, after which there is a very rapid increase. A dotted line shows peak warming at +3°C by 2100 if currrent policies continue.
November 6, 2025 at 3:26 PM
marcrr.bsky.social
Reminds me of this @xkcd.com
xkcd.com/2456
Various joke article titles. Full transcript at https://www.explainxkcd.com/wiki/index.php/2456:_Types_of_Scientific_Paper
June 18, 2025 at 7:15 PM
marcrr.bsky.social
Yep www.smbc-comics.com/comic/crack
@zachweinersmith.bsky.social
Cartoon picture. On the left, three robots, on the right, a man working with a tool and sweating. One robot asks Why. Why are they like this? another answers They had a long abusive relationship with evolution.
May 28, 2025 at 9:15 AM
marcrr.bsky.social
Did I mention that I'm recruiting a postdoc? 😁
sifted.eu/articles/us-...
Map of European OECD countries colored according to average annual salary for a postdoc researcher. Switzerland has the most intense color, indicating the highest salary.
May 22, 2025 at 3:50 PM
marcrr.bsky.social
Tweets of @marc_rr on Jan 3, 2020: OK help, this is true, the income GWAS was performed under approval for different research. The approval: https://ukbiobank.ac.uk/2015/06/dr-catherine-gale-university-of-edinburgh/ The word 'income' is absent from the application. 😱 Quote-tweeting @StevePittelli I recently wrote to the UK Biobank (haven’t heard back, yet) re: the study below and requested that the authors of that study be prevented from utilizing the UK Biobank in any future studies. My letter here: https://bit.ly/36eRGUc Thread… 1/ Next tweets: At the end of the application, there is "will investigate whether other characteristics, such as lifestyle, socio-demographic, physical, behavioural or biological factors, help to explain any links between cognitive function and emotions and these health outcomes" That's pretty far from a main focus on genetics of income. Tweets of August 30, 2019, by @marc_rr For @uk_biobank, the broad consent is to "improve the prevention, diagnosis and treatment of illness and the promotion of health". How the hell can this study be justified under such a consent agreement? Screenshot of Broad Institute: In the case of UK Biobank, research subjects submitting samples are “broadly consented." This means subjects consent to having their samples used “to improve the prevention, diagnosis and treatment of illness and the promotion of health throughout society." In the case of the Ganna et al. study, it is debatable as to whether LGBTQIA+-identifying subjects who submitted samples and data to the UK Biobank would have consented if they had been specifically informed that their genetic information would be used for a study like this. Next tweets: I went to check the consent form, in case the opinion was a simplification. No. It's exactly that. At every page, the consent forms speak about health and medical research. So either homosexuality is estimated by the scientists and administrators involved to be an illness, and WTF, or just WTF?
April 15, 2025 at 11:00 AM
marcrr.bsky.social
Finally, we show with a simple evolutionary model that an excess of intermediate pleiotropy for adaptive substitutions is expected following an ecological shift.
Histograms of mutation distributions over pleiotropy. Left column, SSWM regime, right column Greedy walk regime. Top row, all mutations, skewed to low pleiotropiy. Bottom row, fixed mutations, skewed to intermediate pleiotropy under the Greedy walk regime.
April 8, 2025 at 1:08 PM
marcrr.bsky.social
In addition, we had already shown that the convergent genes are expressed in a diversity of cell types in zebrafish development.
Left, clustering over developmental time of convergent gene single-cell expression in zebrafish embryos, showing a set of genes with time-restricted expression. Right, UMAP of the single-cell at 24hpf, 5dpf and 10dpf, with 5 genes highlighted, showing that each is expressed in diverse cell types.
April 8, 2025 at 1:08 PM
marcrr.bsky.social
3- Unique phenotype terms affected by genetic perturbations from @zfinmod.bsky.social: a gene whose KO or knock-down leads to a more complex phenotype, or to more phenotypes described by different terms, is more pleiotropic. Convergent genes again have an excess of intermediate values.
Top left, circos plot of relation between genes and phenotypes affected. Top right, density plot of log2(number of unique phenotype terms) convergent vs. non-convergent genes, showing an excess of intermediate values for convergent genes. Bottom, histogram of number of unique phenotypic terms associated with genetic perturbation per gene.
April 8, 2025 at 1:07 PM
marcrr.bsky.social
2- Tissue-specificity over organs: a gene expressed in more organs is expected to impact more functions, thus be more pleiotropic. Convergent genes have excess intermediate tissue-specificity, whereas most genes have low (housekeeping) or high (organ-specific) values.
Density plots of tissues specificity tau in 11 fish species, for convergent vs. non-convergent genes, showing that in 10 out of 11 species the distribution is unimodal for convergent genes but bimodal for non-covergent genes.
April 8, 2025 at 1:06 PM
marcrr.bsky.social
1- Diversity of @geneontology.bsky.social annotations: a gene with more diverse GO annotations is involved in more functions, thus more pleiotropic. Convergent genes have excess intermediate semantic similarity of their GO annotations.
Density plots of log2(Wang's semantic similarity) of GO terms for convergent vs. non-convergent genes, showing an excess of intermediate values for convergent genes in each of the three GO ontologies: biological process, molecular function, and cellular component. In all cases, KS test p < 2e-16.
April 8, 2025 at 1:05 PM
marcrr.bsky.social
First, we confirm that convergent substitutions in fish coding genes have functional impacts, with more negative BLOSUM scores and more changes in physicochemical properties of amino acids.
Left, boxplot of BLOSUM62 scores of convergent vs non-convergent substitutions, showing lower scores for convergent substitutions; Wilcoxon test p = 6e-7 Right, barplot of 17 amino acid properties for convergent vs non-convergent substitutions, showing larger changes for convergent substitutions for every property
April 8, 2025 at 1:05 PM
marcrr.bsky.social
I have signed the petition "EPFL community demands a competent Ethical Committee with International Law experts". I also attach a screenshot of my comment (not sure why I wrote it in French but too late to change)
www.leslignesbougent.org/petitions/ep...
Marc Ro*********i En tant que professeur à l'UNIL qui a des collaborations avec des universitaires internationaux y compris israéliens, je trouve que les mesures proposées par le rapport UNIL apportent une clarté et un cadre nécessaires et appropriés. Je soutiens les efforts pour mettre en place des cadres similaires dans d'autres universités. Je note que le cadre à l'UNIL ne se focalise pas sur Israël ou un autre conflit, mais est général et sera adaptable à des futures situations auxquelles nous seront confrontés. 09/03/2025, 11:31:32
March 9, 2025 at 10:37 AM
marcrr.bsky.social
This pattern appears driven both by the evolution of developmental processes intrinsic to teeth and by negative and positive selection essentially extrinsic to the teeth, via the level of gene pleiotropy.
January 27, 2025 at 10:23 AM
marcrr.bsky.social
Interestingly, while highly expressed genes are frequent both in early and late molar development, broadly expressed genes are frequent early but not late. Thus early and late conservation have different causes.
Genes expressed at different stages of development harbor different levels of expression and of cell expression specificity in molar. A-D : Genes were split in quantiles according to their average expression level. The proportion of genes associated with each quantile is drawn for each developmental stage and for each molar. Quantile boundaries are computed separately for each molar, respectively quantile 1 to 4 in TPM for mus mx: 1.7, 20.9, 74.9,; for mus md: 1.6, 20.9, 74.8 for ham mx: 1.6, 17.2, 64.2; for ham md:1.5, 17.1 ,63.9. E-H: Genes were split in quantiles according to their pleiotropy. The proportion of genes associated with each quantile is drawn for each developmental stage and for each molar. Quantile boundaries were computed on tau values measured on mouse molar single cell transcriptome, with three threshold values ranging from low to high pleiotropy (25% 0.37, 50% 0.61, 75% 0.85).
January 27, 2025 at 10:23 AM
marcrr.bsky.social
We found a similar pattern in selection on protein coding genes, with lower selective pressure (dN/dS) and higher positive selection (branch-site test) in middle development of molars.
Selective pressure acting on gene coding sequences for each coexpressed gene cluster. Average pairwise dN/dS between mouse and hamster were computed as mean(dN)/mean(dS). Observed values are represented by a black line, intervals show the 95 percentiles obtained by 1000 bootstrap resamplings. <fig 4> Positive selection for genes with expression peaks at different stages of development, in four branches of the rodent tree. Each plot shows the fourth root of the likelihood ratio (ΔlnL) of the models with and without positive selection, taken from Selectome database. Average for coexpressed gene clusters associated with each stage of development are represented. Confidence intervals obtained from 1000 bootstraps. Branches used to compute positive selection are indicated on the rodent tree in (A), and correspond respectively to (B): Muridae, (C): Murinae, (D): Mus, (E): Cricetinae. Gene clusters correspond to mouse upper molar (B-D) or to hamster upper molar (E). Species tree in panel A pruned from Ensembl Compara.
January 27, 2025 at 10:22 AM
marcrr.bsky.social
As expected under coevolution upper and lower molars of the same species have very similar gene expression.
As expected from rapid molar evolution, mouse and hamster gene have rather different gene expression.
More surprising the coexpression clusters are most divergent in middle molar development.
Overlap of expression gene clusters between species. Each cluster corresponds to a set of genes with an expression peak at a specific stage of development in a given tooth and species. Number of genes shared between hamster and mouse coexpression clusters, expressed as a fold enrichment relative to equally sized random clusters, for the upper (A) and the lower (B) molars. Clusters are ordered according to their temporal profiles. Proportion of genes shared between hamster and mouse coexpression clusters peaking at the same stage in both species, in the upper (C) and lower (D) molars. Intervals show the 95 percentiles obtained by 1000 bootstrap resamplings. Expected values obtained by randomly shuffling genes between clusters (dashed lines).
January 27, 2025 at 10:21 AM
marcrr.bsky.social
We obtained RNAseq from tooth germs over the embryonic and postnatal period where the major events of morphogenesis occur, from bud, to cap, to bell stage until differentiation and enamel/dentin secretion, and obtained clusters of coexpressed genes per stage.
Global expression profiles and clusters of co-expressed genes with an expression peak at a specific stage of dental development. (A) PCA based on 11,342 1:1 orthologs across mouse and hamster. Samples are colored per morphological stage, shapes correspond to upper and lower molar samples. (B) Coexpression clusters defined by soft clustering of expression values of each species and tooth type. Each cluster corresponds to a set of genes with an expression peak at a specific stage of dental development. The average profiles of each cluster are represented, with expression levels centered at the value of the first stage and error bars showing standard deviations. Clusters are ordered according to their temporal profiles. Distribution of cluster sizes is given. Genes not associated with a cluster are not represented. (C) Summary of the GO enrichment analysis for the mouse lower molar coexpression clusters; enriched GO terms were categorized as shown in Fig S2, a maximum of 5 representative terms are shown with their fold enrichments.
January 27, 2025 at 10:21 AM
marcrr.bsky.social
We are excited to share our work on molar #EvoDevo 🦷 in mouse and hamster 🐁, led by the amazing team of Marie Sémon and Sophie Pantalacci. We find an inverted hourglass pattern, with a maximum of evolutionary divergence at the bell stage.
www.biorxiv.org/content/10.1...
@lbmcinlyon.bsky.social
Top: hourglass pattern of increasing then decreasing divergence over molar development, with the stages BUD, CAP, BEL (maximum divergence), DIF and SEC. Bottom: illustration of gene processes influencing this pattern, from left to right: image of embryo with “shared: limbs, scales, glands; lung, kidney”; “immunity genes (influx of blood)”; image of embryo and of tooth with “shared: secretion; intrinsic: mineralisation”.
January 27, 2025 at 10:20 AM
marcrr.bsky.social
New (short) preprint, about our BeeBiome portal:
ecoevorxiv.org/repository/v...
BeeBiome logo: a simplified bee picture
January 20, 2025 at 4:26 PM
marcrr.bsky.social
No gods, no masters, no scientific heroes
screenshot of X (Twitter) where I wrote in August 2021: No gods, no masters, no scientific heroes quoting a tweet of Florian Fisch who was quoting David Gorski: "In the last year, my disappointment with Prof. Ioannidis has gotten to the point where it’s hard for me to avoid lumping him with the COVID-19 minimizers/deniers"
January 16, 2025 at 1:59 PM
marcrr.bsky.social
Interesting discussion of causes and consequences of a decline in fieldwork-based research and education in ecology (also relevant to evolution)
doi.org/10.1016/j.tr...
Top box: drivers of decline in fieldwork Arrows linking this through loss capability, loss of opportunity, loss of motivation, to the decline then to a bottom box Bottom box: consequences of decline in fieldwork Description from article: Various social (e.g., societal expectations, regulations), economic (e.g., limited funding, rising research costs), environmental (e.g., ecological degradation, climate change), technological (e.g., use of remote sensing over field observation), educational (e.g., declining focus on natural sciences in school curricula), and cultural (e.g., shifting academic values regarding fieldwork-based studies) factors can reduce the capability, opportunity, and motivation to conduct fieldwork-based research and education. Decreased fieldwork can negatively impact ecological research and education, which in turn can hamper efforts to address the ongoing global biodiversity decline. Several feedback loops likely exist, where declines in fieldwork lead to further reductions in these activities.
January 13, 2025 at 10:15 AM
marcrr.bsky.social
New group photo bioinfo.unil.ch @dee-unil.bsky.social
Photo of lab members standing on the stairs of the natural history museum of Lausanne
December 20, 2024 at 8:53 AM
marcrr.bsky.social
Je suis retombé sur cette BD de @bouletcorp.bsky.social, et ça m'a irrésistiblement fait penser à la génération "d'art" AI, entraîné sur de l'art existant et demain sur des résultats AI
source: bouletcorp.com/rogatons/202...
extrait de BD de Boulet montrant son interactions avec un prof de dessin, qui lui conseille de s'inspirer du réel plutôt que de la BD pour progresser en dessin
December 18, 2024 at 12:25 PM
marcrr.bsky.social
New @dee-unil.bsky.social mug has arrived
Photo of a blue mug in a coffee machine. The mug has the logo of the DEE | Department of ecology and evolution, which are the mountain range visible from the department building.
December 3, 2024 at 9:23 AM
marcrr.bsky.social
This suggests that gene expression variability in fishes can be driven by organ-specific selection on the cis-regulatory sequences. Notably brain-biased genes seem to have evolved under strong purifying selection, not only at the protein-coding sequence level, but also at the cis-regulatory level.
Photos of a gar, a zebrafish and a pike
November 28, 2024 at 11:03 AM