Biology subjects

Dominissini, D.

Publications and source records attributed to Dominissini, D..

7 recordsLinked to original sources

bioRxiv2026

DNT: Diploid Genomic Foundation Model

Clinical interpretation of genetic variation depends on the diploid genotype, including zygosity, allele dosage and whether multiple variants occur in cis on the same homologue or in trans on different homologues. Most genomic language models process haploid sequences or combine independently encoded haplotypes downstream, so they do not directly represent the paired genotype in a single sequence. We introduce a reference-aligned diploid encoding for single-nucleotide variants (SNVs) and short insertions and deletions (indels), together with unphased and phase-retaining tokenizers that accept phased genotypes and convert them to single-sequence diploid representation. Using Nucleotide Transformer v3 backbones, we continue training 8-million- and 100-million-parameter models and evaluate an auxiliary Contrastive Phase Loss (CPL) designed to retain the phasing information of the variants in contextual representations. We evaluate on a novel compound-heterozygous benchmark containing 9,460 examples. Models whose inputs did not distinguish relative phase remained near chance, whereas our diploidic models improved discrimination with AUROC 0.649, compared to 0.506 for the vocabulary-adapted control. These findings establish a method for making diploid genotype information accessible to genomic language models, rather than a universal improvement in variant prediction; validation in naturally observed, accurately phased clinical cohorts remains necessary.

bioinformatics↗

bioRxiv2026

Context-dependent determinants of CRISPR-Cas9 editing efficiency revealed through cross-species endogenous editing analysis

Accurate prediction of CRISPR-Cas9 guide RNA (gRNA) editing efficiency remains limited, particularly outside human systems, where models trained on exogenous human datasets show poor generalization. We analyzed Cas9 efficiency and repair outcomes using novel endogenous editing data from four human cell types, two tomato cell types, and cells from giant river prawn and black soldier fly. While integrating publicly available predictors via ensemble frameworks improved performance, our analysis revealed hundreds of novel features affecting activity. Crucially, dominant features related to sites competition for gRNA, and local geometric properties varied across systems, highlighting the strong context dependence of Cas9 efficiency and arguing against a universal model. Interestingly, codon usage bias-based features also emerged as informative predictors, as they are proxies for chromatin accessibility. In contrast, trends in repair outcomes remained conserved. This work provides essential resources for more generalizable CRISPR guide design.

synthetic biology↗

bioRxiv2026

GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models

The rapid scaling of Genomic Foundation Models (GFMs) has created a critical need for standardized evaluation frameworks. Current benchmarking practices are often fragmented, relying on model-specific preprocessing and inconsistent metric implementations that hinder reproducible comparisons. We present GFMBench-API, a high-level Python interface designed to unify the evaluation lifecycle of GFMs. GFMBench-API provides a modular "middleware" architecture that decouples model-specific tokenization and embedding logic from task-specific data streams and performance metrics. By standardizing the input/output schemas for common genomic tasks, such as regulatory element prediction, variant effect scoring, and long-range interaction mapping, GFMBench-API enables researchers to integrate new models or tasks with minimal "glue code." Our interface ensures mathematical consistency across evaluations, providing a robust foundation for the transparent and systematic benchmarking of GFMs.

genomics↗

bioRxiv2025

i6A-seq maps N6-isopentenyladenosine and uncovers its role as a regulator of mRNA stability through recruitment of DIS3L2

Dynamic post-transcriptional RNA modifications are crucial regulators of RNA metabolism and cell fate. Recent advances in sequencing, combined with antibody, enzymatic, or chemical approaches, have enabled transcriptome-wide mapping of these modifications. While prevalent in tRNAs and rRNAs, a growing number of modifications also adorn lower-abundance mRNA transcripts. Global mapping efforts, particularly for N6-methyladenosine (m6A), uncovered the affected mechanisms governing RNA metabolism, shedding light on novel, modification-dependent, modes of gene regulation. N6-isopentenyladenosine (i6A) is a conserved tRNA modification involved in translation fidelity and efficiency. i6A depletion is linked to mitochondrial defects and human diseases. Current i6A detection methods are low-throughput, and its lack of base-pairing effects makes sequencing-based identification challenging. Here we developed a novel and robust mapping technique, i6A-seq, a method that utilizes antibody-mediated enrichment and iodine chemical labeling to generate a reverse transcription signature, enabling transcriptome-wide i6A detection. Our global mapping revealed hundreds of i6A sites in human and mouse mRNA, indicating its presence in this RNA species, with conserved isopentenylome features. These sites exhibit a typical consensus sequence, primarily in coding transcripts (CDS), preferentially within lysine codons. Consistent with its role in tRNA, tRNA-isopentenyltransferase (TRIT1) also appears to install i6A in mRNA. Manipulation of TRIT1 revealed that i6A regulates the expression of a subset of genes through mRNA decay, specifically those isopentenylated at the CDS and translated on ER-bound ribosomes. Importantly, DIS3 like 3-5 exoribonuclease 2 (DIS3L2), an RNA exoribonuclease known to regulate ER-translated mRNA, was identified as the first i6A reader protein. Our findings introduce i6A as a new modified nucleotide that decorates mRNA, allowing us to decipher its regulatory roles in gene expression. This study not only establishes the presence and location of i6A in mRNA, but also uncovers its first functional mechanism. Similar to the knowledge accumulated on m6A, this work paves the way for further discoveries with potential relevance for understanding gene regulation, disease diagnosis, and therapy.

molecular biology↗

bioRxiv2023

Small Extracellular Vesicles from Failing Heart Accelerate Tumor Growth

BackgroundMyocardial infarction (MI) and heart failure (HF) are associated with an increased incidence of cancer. The mechanism is complex and unclear. Here, we aimed to test our hypothesis that cardiac small extracellular vesicles (sEVs), particularly cardiac mesenchymal stromal cells-derived sEVs (cMSC-sEVs), contribute to the link between post-MI HF and cancer. MethodsWe purified and characterized sEVs from the whole heart and cultured cMSCs. Then, we analyzed cMSC-EV cargo and pro-neoplastic effects on several types of cancer cell lines, macrophages, and endothelial cells. Next, we modeled post-MI HF along with heterotopic and orthotopic lung and breast cancer tumors in mice. We used cMSC-sEV transfer to assess sEV biodistribution and its effect on tumor growth. Finally, we tested the effects of sEV depletion and spironolactone treatment on cMSC-EV release and tumor growth. ResultsPost-MI hearts, particularly cMSCs, produced more sEVs with pro-neoplastic cargo than non-failing hearts did. Proteomic analysis revealed unique protein profiles and higher quantities of tumor-promoting cytokines, proteins, and microRNAs in cMSC-sEVs from failing hearts. The pro-neoplastic effects of cMSC-sEVs varied with different types of cancer cells, substantially affecting lung cancer cells relative to other more aggressive cancer cell lines. We also found that post-MI cMSC-sEVs activated resting macrophages into pro-angiogenic and pro-tumorigenic states in vitro. At 28-day follow-up analysis, mice with post-MI HF developed larger lung tumors than did sham-MI mice. Adoptive transfer of cMSC-sEVs from failing hearts accelerated lung tumor growth, and biodistribution analysis revealed an accumulating cMSC- sEVs in tumor cells along with accelerated tumor cell proliferation. Significantly, sEV depletion reduced the tumor-promoting effects of HF, and adoptive transfer of cMSC-sEVs from failing hearts partially restored it. Finally, post-MI spironolactone treatment reduced the number of cMSC-sEVs and suppressed tumor growth during post-MI HF. ConclusionsFor the first time, we show that cardiac sEVs, specifically cMSC-sEVs from post-MI failing hearts, carry multiple pro-tumorigenic factors. Uptake of cMSC-sEVs by cancer cells accelerates tumor growth. Post-MI spironolactone treatment reduces the associated tumor growth. Thus, we provide new insight into the link between post-MI HF and cancer and propose a translational option to mitigate this deadly association.

cancer biology↗

bioRxiv2020

Deep and accurate detection of m6A RNA modifications using miCLIP2 and m6Aboost machine learning

N6-methyladenosine (m6A) is the most abundant internal RNA modification in eukaryotic mRNAs and influences many aspects of RNA processing. miCLIP (m6A individual-nucleotide resolution UV crosslinking and immunoprecipitation) is an antibody-based approach to map m6A sites with single-nucleotide resolution. However, due to broad antibody reactivity, reliable identification of m6A sites from miCLIP data remains challenging. Here, we present miCLIP2 in combination with machine learning to significantly improve m6A detection. The optimised miCLIP2 results in high-complexity libraries from less input material. Importantly, we established a robust computational pipeline to tackle the inherent issue of false positives in antibody-based m6A detection. The analyses are calibrated with Mettl3 knockout cells to learn the characteristics of m6A deposition, including m6A sites outside of DRACH motifs. To make our results universally applicable, we trained a machine learning model, m6Aboost, based on the experimental and RNA sequence features. Importantly, m6Aboost allows prediction of genuine m6A sites in miCLIP2 data without filtering for DRACH motifs or the need for Mettl3 depletion. Using m6Aboost, we identify thousands of high-confidence m6A sites in different murine and human cell lines, which provide a rich resource for future analysis. Collectively, our combined experimental and computational methodology greatly improves m6A identification. HighlightsO_LImiCLIP2 produces complex libraries to map m6A RNA modifications C_LIO_LIMettl3 KO miCLIP2 allows to identify Mettl3-dependent RNA modification sites C_LIO_LIMachine learning predicts genuine m6A sites from human and mouse miCLIP2 data without Mettl3 KO C_LIO_LIm6A modifications occur outside of DRACH motifs and associate with alternative splicing C_LI

genomics↗

bioRxiv2017

N6-methyladenosine modification and the YTHDF2 reader protein play cell type specific roles in lytic viral gene expression during Kaposi’s sarcoma-associated herpesvirus infection

Methylation at the N6 position of adenosine (m6A) is a highly prevalent and reversible modification within eukaryotic mRNAs that has been linked to many stages of RNA processing and fate. Recent studies suggest that m6A deposition and proteins involved in the m6A pathway play a diverse set of roles in either restricting or modulating the lifecycles of select viruses. Here, we report that m6A levels are significantly increased in cells infected with the oncogenic human DNA virus Kaposis sarcoma-associated herpesvirus (KSHV). Transcriptome-wide m6A-sequencing of the KSHV-positive renal carcinoma cell line iSLK.219 during lytic reactivation revealed the presence of m6A across multiple kinetic classes of viral transcripts, and a concomitant decrease in m6A levels across much of the host transcriptome. However, we found that depletion of the m6A machinery had differential pro- and anti-viral impacts on viral gene expression depending on the cell-type analyzed. In iSLK.219 and iSLK.BAC16 cells the pathway functioned in a pro-viral manner, as depletion of the m6A writer METTL3 and the reader YTHDF2 significantly impaired virion production. In iSLK.219 cells the defect was linked to their roles in the post-transcriptional accumulation of the major viral lytic transactivator ORF50, which is m6A modified. In contrast, although the ORF50 mRNA was also m6A modified in KSHV infected B cells, ORF50 protein expression was instead increased upon depletion of METTL3, or, to a lesser extent, YTHDF2. These results highlight that the m6A pathway is centrally involved in regulating KSHV gene expression, and underscore how the outcome of this dynamically regulated modification can vary significantly between cell types.\n\nAuthor SummaryIn addition to its roles in regulating cellular RNA fate, methylation at the N6 position of adenosine (m6A) of mRNA has recently emerged as a mechanism for regulating viral infection. While it has been known for over 40 years that the mRNA of nuclear replicating DNA viruses contain m6A, only recently have studies began to examine the distribution of this modification across viral transcripts, as well as characterize its functional impact upon viral lifecycles. Here, we apply m6A-sequencing to map the location of m6A modifications throughout the transcriptome of the oncogenic human DNA virus Kaposis sarcoma-associated herpesvirus (KSHV). We show that the m6A machinery functions in a cell type specific manner to either promote or inhibit KSHV gene expression. Thus, the KSHV lifecycle is impacted by the m6A pathway, but the functional outcome may depend on cell lineage specific differences in m6A-based regulation.

microbiology↗