Biology subjects

Mulder, L.

Publications and source records attributed to Mulder, L..

4 recordsLinked to original sources

bioRxiv2022

Development of an HIV reporter virus that identifies latently infected CD4+ T cells

There is no cure for HIV infection as the virus establishes a latent reservoir, which escapes highly active antiretroviral treatments. One major obstacle is the difficulty to identify cells that harbor latent proviruses. We devised a novel viral vector that carries a series of versatile reporter molecules that are expressed in an LTR-dependent or LTR-independent manner and allows to accurately distinguish productive from latent infection. Using primary human CD4+ T cells, we show that transcriptionally silent proviruses make up over 50% of all infected cells. These latently infected cells harbor proviruses, but lack evidence for viral transcription. LTR silent integrations occurred to variable degrees in all CD4+ T-subsets examined with CD4+ TEM and CD4+ Treg displaying the highest frequency. Viral vectors such as the one described here, permit interrogation HIV latency at a single-cell resolution, revealing mechanisms of latency establishment and allowing for the characterization of effective latency reversing agents.

microbiology↗

bioRxiv2021

Biological Constraints Can Improve Prediction inPrecision Oncology

Many gene signatures have been developed by applying machine learning (ML) on omics profiles, however, their clinical utility is often hindered by limited interpretability and unstable performance in different datasets. Here, we show the importance of embedding prior biological knowledge in the decision rules yielded by ML approaches to build robust classifiers. We tested this by applying different ML algorithms on gene expression data to predict three difficult cancer phenotypes: bladder cancer progression to muscle invasive disease; response to neoadjuvant chemotherapy in triple-negative breast cancer, and prostate cancer metastatic progression. We developed two sets of classifiers: mechanistic, by restricting the training process to features capturing a specific biological mechanism; and agnostic, in which the training didnt use any a priori biological information. Mechanistic models had a similar or better performance to their agnostic counterparts in the testing data, with enhanced stability, robustness, and interpretability. Our findings support the use of biological constraints to develop robust and interpretable gene signatures with high translational potential. MotivationOmics-based gene signatures often suffer from overfitting and reduced performance when tested on independent data. This usually results from the discrepancy between the high number of features compared to the much smaller number of samples used in the training process, which results in the machine learning algorithm perfectly fitting the training data with a subsequent deterioration in performance in independent cohorts. We introduce a mechanistic framework to mitigate overfitting and improve interpretability by constraining the training process to simple rank-based decision rules recapitulating relevant, cancer-related, biological mechanisms. Our approach aims at reducing the number of training variables to a pre-defined set of biologically important features in the form of gene pairs. The classification mechanism depends entirely on the relative ordering of these pairs, making it robust to data preprocessing techniques, improving the overall interpretability of the resulting models with significant translational implications. Most importantly, these pairs are configured in such a way that the decision rules resulting from the genes relative order embed and recapitulate specific biological mechanism, inherently enhancing the classifiers interpretability.

bioinformatics↗

bioRxiv2021

A photo-switchable yeast isocitrate dehydrogenase to control metabolic flux through the citric acid cycle

For various research questions in metabolism, it is highly desirable to have means available, with which the flux through specific pathways can be perturbed dynamically, in a reversible manner, and at a timescale that is consistent with the fast turnover rates of metabolism. Optogenetics, in principle, offers such possibility. Here, we developed an initial version of a photo-switchable isocitrate dehydrogenase (IDH) aimed at controlling the metabolic flux through the citric acid cycle in budding yeast. By inserting a protein-based light switch (LOV2) into computationally identified active/regulatory-coupled sites of IDH and by using in vivo screening in Saccharomyces cerevisiae, we obtained a number of IDH enzymes whose activity can be switched by light. Subsequent in-vivo characterization and optimization resulted in an initial version of photo-switchable (PS) IDH. While further improvements of the enzyme are necessary, our study demonstrates the efficacy of the overall approach from computational design, via in vivo screening and characterization. It also represents one of the first few examples, where optogenetics were used to control the activity of a metabolic enzyme.

synthetic biology↗

bioRxiv2018

Telescope: Characterization of the retrotranscriptome by accurate estimation of transposable element expression

Characterization of Human Endogenous Retrovirus (HERV) expression within the transcriptomic landscape using RNA-seq is complicated by uncertainty in fragment assignment because of sequence similarity. We present Telescope, a computational software tool that provides accurate estimation of transposable element expression (retrotranscriptome) resolved to specific genomic locations. Telescope directly addresses uncertainty in fragment assignment by reassigning ambiguously mapped fragments to the most probable source transcript as determined within a Bayesian statistical model. We demonstrate the utility of our approach through single locus analysis of HERV expression in 13 ENCODE cell types. When examined at this resolution, we find that the magnitude and breadth of the retrotranscriptome can be vastly different among cell types. Furthermore, our approach is robust to differences in sequencing technology, and demonstrates that the retrotranscriptome has potential to be used for cell type identification. Telescope performs highly accurate quantification of the retrotranscriptomic landscape in RNA-seq experiments, revealing a differential complexity in the transposable element biology of complex systems not previously observed. Telescope is available at github.com/mlbendall/telescope.\n\nAuthor SummaryAlmost half of the human genome is composed of Transposable elements (TEs), but their contribution to the transcriptome, their cell-type specific expression patterns, and their role in disease remains poorly understood. Recent studies have found many elements to be actively expressed and involved in key cellular processes. For example, human endogenous retroviruses (HERVs) are reported to be involved in human embryonic stem cell differentiation. Discovering which exact HERVs are differentially expressed in RNA-seq data would be a major advance in understanding such processes. However, because HERVs have a high level of sequence similarity it is hard to identify which exact HERV is differentially expressed. To solve this problem, we developed a computer program which addressed uncertainty in fragment assignment by reassigning ambiguously mapped fragments to the most probable source transcript as determined within a Bayesian statistical model. We call this program, \"Telescope\". We then used Telescope to identify HERV expression in 13 well-studied cell types from the ENCODE consortium and found that different cell types could be characterized by enrichment for different HERV families, and for locus specific expression. We also showed that Telescope performed better than other methods currently used to determine TE expression. The use of this computational tool to examine new and existing RNA-seq data sets may lead to new understanding of the roles of TEs in health and disease.

genomics↗