Dante DeAscanis’ Post

🧐 Do you remember to check the underlying assumptions of an analytical tool? In my latest paper, you can get a glimpse of the pitfalls that can occur when some sneaky eukaryotic or archaeal DNA gets mixed into bacterial datasets.  🌲 Phylogenetic trees should be inspected in bacterial 16S rRNA datasets ✖️ Misclassified sequences in the phylogenetic tree can cause dramatic changes in community assembly analysis 🌍 Manipulating the metacommunity richness had little effect on the conclusions from the analytical tools applied. However, how to define a cut-off in metacommunity membership remains an open question.  📊 Understanding the assumptions of the methods used is important to avoid false conclusions! Happy to discuss! https://lnkd.in/dGecQPpa

We can compare the different types of grapes used to produce a certain wine with a box of crayons 🖍 Now imagine that the produced wine is a picture drawn with those crayons. We can say that the authors of this paper developed a methodology that allow to figure out what specific crayons were used to paint the draw even if the colors are all mixed up. Got curious? Know more about this investigation: https://bioisi.pt/news/82 Read the full paper first-authored by Sara Barrias, PhD Student at the DNA & RNA Sensing Lab (at Universidade de Trás-os-Montes e Alto Douro ), coordinated by Paula Martins-Lopes, group leader of the same Lab, and published in the Food Control Journal: https://bit.ly/4gbw0fU #BioISiDigest #Wineauthenticity #FoodControl #DNAsensing

Weds Mar. 6, 2024 at 11:00 AM ET is the next meeting of a Special Interest Group, "Modeling Ensembles and Alternative Conformations of Proteins: Critical evaluation of (protein) structure prediction (CASP)." Every 2 years, this group evaluates computational protein structure prediction algorithms in a blind competition. This year, they have a special entry for data from the Bruce Donald Lab and the Terry Oas Lab which allows the inference of a continuous distribution of protein structures rather than a rigid model. The Mar. 6 meeting is charged with defining the parameters of the competition, what data we will give them, and how the contestant entries will be evaluated. Zoom link: https://lnkd.in/dtdGspjt

Can we classify objects at a finer granularity than the current labels allow? Our ICML paper, "Fine-grained Classes and How to Find Them" has the answer. Coarse labels often arise due to a lack of expert knowledge or budget constraints. We show that these broad labeling strategies can be broken down into a two-step procedure: fine-grained labeling followed by fine-to-coarse class remapping. The two-step procedure can be effectively modeled using a deep classifier and an undirected bipartite graph. Interestingly, joint training of the two components alternates between gradient-based optimization and discrete optimization of integer programs. When applied to animal images, our method discovers meaningful subspecies. Similarly, in single-cell transcriptomics data, we uncover meaningful cell subtypes. See you in Vienna next month! Project page: https://lnkd.in/dsrNuyUU Paper: https://lnkd.in/drBUrNd3 Code: https://lnkd.in/drkZpca9

Know thy intrinsic gravitation in life science through graphical representation 92 Equivalence in working principles of intrinsic gravitation in stellar mass and formation of organelles in bio-cell.

Soluble membrane protein analogs that are structured, folded and functional shown with GPCRs and Claudins! If extended to the vast membrane protein proteome this pipeline may open the door to a much easier way to study these targets.

I'll be presenting this poster in a few weeks as part of a week-long event organised by SFI Centre for Research Training in Foundations of Data Science , based on the most recent paper by myself and my supervisor, Marius Ghergu. It concerns the Gierer-Meinhardt system of PDEs, conceived to model morphogenesis, the biochemical process of cell-shaping. We break new ground in the setting of an "exterior domain", i.e. the entire real space bar with a "hole" (compact set) punched through it. Results include conditions for nonexistence and existence, as well as the asymptotic behaviour of solutions in the latter case. For the full paper, see ArXiv. https://lnkd.in/eM_eVgCn

Are you curious how the Spring Platform has been used by researchers to push their science forward? Watch this short video featuring Spring's Rachel Jacobson to learn how the Spring Engine was used for inflammasome screening and profiling

All types of analytical to cellular research are done here.......................

Read the article: https://lnkd.in/d-VTGtVa In the evolving landscape of scientific research, recent paper retractions due to western blot image manipulation have cast a spotlight on the critical need for robust data integrity and stringent manuscript approval processes. At its core, data integrity ensures that the data presented in scientific papers accurately reflect experimental reality. Therefore, it is imperative for journals to enforce stricter prepublication requirements and for researchers to follow best practices for data normalization, particularly in western blotting, to safeguard the accuracy of published data. In this article in Lab Manager, Chelsea Pratt from Bio-Rad Laboratories discusses the importance of data normalization and how stain-free western blotting workflow can help scientists maintain data integrity whilst adhering to new publication guidelines. Learn more about stain-free western blotting ➡️ https://lnkd.in/gzTfRWVV #WesternBlotting #DataNormalization #LifeScience