Biology subjects

Bours, V.

Publications and source records attributed to Bours, V..

4 recordsLinked to original sources

bioRxiv2023

Assessment of the normal cell contamination impact on tumour samples analysed with SNP arrays: The signal confusion nightmare.

Recent advances in high-throughput technologies enable a more comprehensive interpretation of the tumour evolution through the study of the intra-tumour heterogeneity. Several algorithms, however, often relies on the use of models that described the top of the iceberg regarding the stromal contamination of the samples, making diagnosis difficult to assess. Indeed, such as radio wave receivers, tools to analyse high-throughput technologies data, are used to enable the discrimination between multiple signals differing in frequencies. However, such tools often look at the average frequency more than distinct signals, leading to analyse a confused signal. This confusion could dramatically lead to a mis-interpretation of the real data, especially during the diagnosis as it relies on the choice of a unique scenario among many others. Here, we describe how this signal confusion occurs in the most classical DNA microarray analysis of tumours and we provide statistics to determine how many other possible scenario can lead the same signals, in order to improve the robustness of pigeon hole logic based analysis. Based on simulations, where a unique tumour population was diluted by an increasing gradient of normal cells, we underline the causes and consequences of such signal confusion for up to five allelic copies. Despite the removal of all technical biaises and background noise, we show how the signal confusion remains systematically present in the commonly used DNA microarray analysis, especially for the genotypes AAAAB, AAAB and AAB for copy numbers 5, 4 and 3 respectively, as well as their symmetric combinations for the B allele.

bioinformatics↗

bioRxiv2023

Assessing Random Forest self-reproducibility for optimal short biomarker signature discovery

Biomarker signature discovery remains the main path to develop clinical diagnostic tools when the biological knowledge on a pathology is weak. Shortest signatures are often preferred to reduce the cost of the diagnostic. The ability to find the best and shortest signature relies on the robustness of the models that can be built on such set of molecules. The classification algorithm that will be used is selected based on the average performance of its models, often expressed via the average AUC. However, it is not garanteed that an algorithm with a large AUC distribution will keep a stable performance when facing data. Here, we propose two AUC-derived hyper-stability scores, the HRS and the HSS, as complementary metrics to the average AUC, that should bring confidence in the choice for the best classification algorithm. To emphasize the importance of these scores, we compared 15 different Random Forests implementation. Additionally, the modelization time of each implementation was computed to further help deciding the best strategy. Our findings show that the Random Forest implementation should be chosen according to the data at hand and the classification question being evaluated. No Random Forest implementation can be used universally for any classification and on any dataset. Each of them should be tested for both their average AUC performance and AUC-derived stability, prior to analysis. Author summaryTo better measure the performance of a Machine Learning (ML) implementation, we introduce a new metric, the AUC hyper-stability, to be used in parallel with the average AUC. This AUC hyper-stability is able to discriminate ML implementations that show the same AUC performance. This metric can therefore help researchers in choosing the best ML method to get stable short predictive biomarker signatures. More specifically, we advocate a tradeoff between the average AUC performance, the hyper-stability scores, and the modeling time.

bioinformatics↗

bioRxiv2020

A phylodynamic workflow to rapidly gain insights into the dispersal history and dynamics of SARS-CoV-2 lineages

Since the start of the COVID-19 pandemic, an unprecedented number of genomic sequences of the causative virus (SARS-CoV-2) have been generated and shared with the scientific community. The unparalleled volume of available genetic data presents a unique opportunity to gain real-time insights into the virus transmission during the pandemic, but also a daunting computational hurdle if analysed with gold-standard phylogeographic approaches. We here describe and apply an analytical pipeline that is a compromise between fast and rigorous analytical steps. As a proof of concept, we focus on the Belgium epidemic, with one of the highest spatial density of available SARS-CoV-2 genomes. At the global scale, our analyses confirm the importance of external introduction events in establishing multiple transmission chains in the country. At the country scale, our spatially-explicit phylogeographic analyses highlight that the national lockdown had a relatively low impact on both the lineage dispersal velocity and the long-distance dispersal events within Belgium. Our pipeline has the potential to be quickly applied to other countries or regions, with key benefits in complementing epidemiological analyses in assessing the impact of intervention measures or their progressive easement.

evolutionary biology↗

bioRxiv2017

Transcriptome wide analysis of natural antisense transcripts shows their potential role in breast cancer.

Non-coding RNAs (ncRNA) represent at least 1/5 of the mammalian transcript amount, and about 90% of the genome length is actively transcribed. Many ncRNAs have been demonstrated to play a role in cancer. Among them, natural antisense transcripts (NAT) are RNA sequences which are complementary and overlapping to those of protein-coding transcripts (PCT). NATs were punctually described as regulating gene expression, and are expected to act more frequently in cis than other ncRNAs that commonly function in trans. In this work, 22 breast cancers expressing estrogen receptors and their paired healthy tissues were analyzed by strand-specific RNA sequencing. To highlight the potential role of NATs in gene regulations occurring in breast cancer, three different gene extraction methods were used: differential expression analysis of NATs between tumor and healthy tissues, differential correlation analysis of paired NAT/PCT between tumor and healthy tissues, and NAT/PCT read count ratio variation between tumor and healthy tissues. Each of these methods yielded lists of NAT/PCT pairs that were demonstrated to be enriched in survival-associated genes on an independent cohort (TCGA). This work allows to highlight NAT lists that display a strong potential to affect the expression of genes involved in the breast cancer pathology.

genomics↗