Explore the vast range of titles available.

Circulating microRNAs are non-invasive biomarkers that can be used for breast cancer diagnosis. However, differences in cancer tissue microRNA expression are observed in populations with different genetic/environmental backgrounds. This work aims at checking if a previously identified diagnostic circulating microRNA signature is efficient in other genetic and environmental contexts, and if a universal circulating signature might be possible. Two populations are used: women recruited in Belgium and Rwanda. Breast cancer patients and healthy controls were recruited in both populations (Belgium: 143 primary breast cancers and 136 healthy controls; Rwanda: 82 primary breast cancers and 73 healthy controls; Ntot = 434), and cohorts with matched age and cancer subtypes were compared. Plasmatic microRNA profiling was performed by RT-qPCR. Random Forest was used to (1) evaluate the performances of the previously described breast cancer diagnostic tool identified in Belgian-recruited cohorts on Rwandan-recruited cohorts and vice versa; (2) define new diagnostic signatures common to both recruitment sites; (3) define new diagnostic signatures efficient in the Rwandan population. None of the circulating microRNA signatures identified is accurate enough to be used as a diagnostic test in both populations. However, accurate circulating microRNA signatures can be found for each specific population, when taken separately.

Long non-coding (lnc)RNAs have been shown to have central roles in stress responses, cell identity and developmental processes in multicellular organisms as well as in unicellular fungi. Previous works have shown the occurrence of lncRNAs in diatoms, namely in Phaeodactylum tricornutum , many of which being expressed under specific stress conditions. Interestingly, P. tricornutum is the only known diatom that has a demonstrated morphological plasticity, occurring in three distinct morphotypes: fusiform, triradiate and oval. Although the morphotypes are interchangeable, the fusiform is the dominant one while both the triradiate and the oval forms are less common, the latter often being associated with stress conditions such as low salinity and solid culture media, amongst others. Nonetheless, the molecular basis underpinning morphotype identity in P. tricornutum remains elusive. Using twelve previously published transcriptomic datasets originating from the three morphotypes of P. tricornutum , we sought to investigate the expression patterns of lncRNAs (lincRNAs and NATs) in these distinct morphotypes, using pairwise comparisons, in order to explore the putative involvement of these noncoding molecules in morphotype identity. We found that differentially expressed lncRNAs cluster according to morphotype, indicating that lncRNAs are not randomly expressed, but rather seem to provide a specific (noncoding) transcriptomic signature of the morphotype. We also present evidence to suggest that the major differences in DE genes (both noncoding and coding) between the stress related oval morphotype and the most common fusiform morphotype could be due, to a large extent, to the hyposaline culture conditions rather than to the morphotype itself. However, several lncRNAs associated to each one of the three morphotypes were identified, which could have a potential role in morphotype (or cell) identity in P. tricornutum , similar to what has been found in both animals and plant development.

Breast cancer is a complex disease, and it is the most common cause of morbidity and mortality among women worldwide. In Sub-Saharan Africa, the clinical characteristics and tumor profiles of breast cancer are still unknown. In the present study we aimed to determine breast tumor profiles of the Rwandan patients in relation to age and tumor stages. We compare our findings to related results from other sub-Saharan Africa studies. Data on age at diagnosis, tumor stage, and hormonal profiles of 138 patients diagnosed between January 2015 and December 2018 were retrospectively retrieved from electronic medical records at three referral hospitals in Rwanda. We compared our results to related findings reported in other Sub-Saharan African countries. All statistical analyses were done using SPSS Inc., Chicago, IL, USA, version 20 and R software languages. The mean age at diagnosis was 49.7 years (SD = 13) and ranged from 17 to 86 years. The majority of patients (57.2%) were diagnosed before 50 years of age compared with 42.8% aged > 50 years. Tumor stage III was the commonest accounting for 62% followed by stage II with 24.8%. The distribution of breast tumor subtypes was ER−, PR−, HER2−: 37.7%; ER+, PR+, HER2−: 31.2%; ER−, PR−, HER2+: 14.5%; ER+, PR+, HER2+: 5.1%; and other subtypes represented 11.6%. There was no statistically significant difference in age and tumor stages between the molecular subtypes. Our findings revealed the predominance of hormonal negative tumors among Rwandan patients with breast cancer. Triple negative was found to be the most common breast tumor subtype regardless of age and tumor stage. Larger prospective studies could examine genetics and environmental factors that may play a role in the differences of tumor characteristics in Sub-Saharan populations.

Abstract Historically neglected or considered to be mere transcriptional noise, long non-coding RNAs (lncRNAs) are now emerging as central, regulatory molecules in a multitude of eukaryotic species, from animals to plants to fungi. Yet, our knowledge about the occurrence of these molecules in the marine environment is still elusive. To help fill this knowledge gap, we have developed LncPlankton, a comprehensive database of candidate marine lncRNAs. By integrating the predictions derived from 10 distinctive coding potential prediction tools in a majority voting setting, we have identified over 2M potential lncRNAs distributed across 414 marine plankton species from over nine different phyla. A user-friendly, open-access web interface of the database has been implemented to facilitate exploration (https://www.lncplankton.bio.ens.psl.eu/). We believe LncPlankton will serve as a rich resource for studies of lncRNAs, which will contribute to small- and large-scale analyses in a wide range of marine plankton species and allow comparative studies between them and well beyond the marine environment.

Long considered as transcriptional noise, long non-coding RNAs (lncRNAs) are emerging as central, regulatory molecules in a multitude of eukaryotic species, from plants to animals to fungi. Yet, our knowledge about the occurrence of these molecules in the marine environment, namely in planktonic protists, is still elusive. To fill this gap of knowledge we developed LncPlankton v1.0, which is the first comprehensive database of marine plankton lncRNAs. By integrating the predictions derived from ten distinctive coding potential prediction tools in a majority voting setting, we identified 2,210,359 lncRNAs distributed across 414 marine plankton species from over nine different phyla. A user friendly, open-access web interface for the exploration of the database was implemented (https://www.lncplankton.bio.ens.psl.eu/). We believe LncPlankton v1.0 will serve as a rich resource for studies of lncRNAs that will contribute to small- and large-scale analyses in a wide range of marine plankton species and allow comparative analysis well beyond the marine environment.Competing Interest StatementThe authors have declared no competing interest.Footnotes* https://www.lncplankton.bio.ens.psl.eu/

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. To 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.