Javier

@JudgeTheZipper Lo hiciste gracias a tu esfuerzo y a una educación pública y de calidad

Hijo de ama de casa y funcionario de grupo B (único ingreso en casa). Cuatro hermanos. Mi madre: estudiad lo que queráis el tiempo que necesitéis, que aquí se ponen las mismas lavadoras y la mismas lentejas para todos. Ahora soy juez y resulta que mis padres eran ricos. Y yo que pensaba que me querían y que su prioridad era nuestra educación.

@JudgeTheZipper Lo hiciste gracias a tu esfuerzo y a una educación pública de calidad.

@AgEInves @CienciaGob @CrueUniversidad @CSIC @SaludISCIII @FECYT_Ciencia @RaquelYotti @DomenecE Thank you @AgEInves for continuing to fund our research program

‼️ La @AgEInves publica la Propuesta de Resolución Provisional de los Proyectos Generación de Conocimiento 2022 💰625 Millones € aei.gob.es/sites/default/… #FondosEuropeos #Ciencia

Do you like method development and want to work on a new generation of cryo-EM sample supports? Join us for a 3-yr postdoctoral fellowship to find fun and supportive colleagues, free (and good) coffee, a great department and state-of-the-art facilities. edu.nl/p738p

¿Ya sabéis con quién vais a ir? ☺️

@theNASciences @TexasScience WoW! Congrats @McLellan_Lab !!

The winner of the 2023 NAS Award in Molecular Biology is Jason S. McLellan of @TexasScience! He is being honored for pioneering work in the molecular and structural biology of viral surface proteins. Learn more: bit.ly/3Xu5GnF #NASaward #MolecularBiology

We’re pleased to announce the 2023 NAS Awards recipients! These 16 individuals are honored for their extraordinary scientific achievements in a wide range of fields spanning the physical, biological, and medical sciences. bit.ly/402ovjs #NASaward #Science #Awards

1) En pocos días la UCM nombrará a Isabel Díaz Ayuso “alumna ilustre”. He evitado hacer pública mi postura sobre dicho nombramiento con la esperanza de que no se llevaría a cabo. Hoy, pérdida toda expectativa, hago públicas las razones por las que discrepo de ese nombramiento.

¿Se puede estar en contra de esto?

Cryo-EM map anisotropy can be attenuated by map post-processing and a new method for its estimation biorxiv.org/cgi/content/sh…

@builab @dynein_awk @TheRealCSBlack @ahmad__khalifa Contrats Huy!

@delarosatrevin Enhorabuena Josemi! Todo lo mejor para tu nueva aventura americana ☺️

I haven't tweeted in a while but...a new stage in my life begins and it might deserve a tweet. For many reasons we decided to leave the comfort in Swedish and start the American adventure. Really excited to join Pirate Fernandez and his crew at the CryoEM facility in St Jude!!!

@IsraelF96135088 DeepEMHancer trains on a large set of different maps and LocScaled maps. I don't see why two specific DE and LS maps should be identical. Also, DeepEMhancer performs masking based on PDB maps. So DeepEMhancer = local scaling + tight masking

I have two questions: -why deepEMhacer and LocScale maps looks different? deepEMhacer is trained on LocScale generated maps no? Why they are different? -deepEMhacer map "looks" better, with RNA base-separation quality in a region with ~4Å resolution. How is this possible?

OK, sequel of the sharpening drama, this time a bit more interesting. Same map as yesterday (~2.2A bacterial ribosome with 2xtRNAs) but this time focused on the small subunit (30S) head, which is always flexible. Unsharpened map colored according to local resolution estimation 👇

@CryoTUD @SjorsScheres @biochem_fan I would have to take a look to the code to refresh but I think the normalization is based on percentil values obtained from the maps ( (x-p5)/(p95-p5)) followed by a clipping process. This will compress the histogram

@1avier @SjorsScheres @biochem_fan Thanks Javier! Agreed, but it would not necessarily need to be restricted to positive numbers depending on the activation function used. Normalisation should not affect the shape of the fall-off though. Could it be due to striding?

I am not OK with refinement against deepEMhancer-ed maps but apparently many people think it is OK. Fine. Maybe I am too old-fashioned.

@SjorsScheres @biochem_fan @IUCrJ I have not read the paper in detail but I think authors claim to use the previous knowledge accumulated in deepEMhancer to improve the alignment and seems to work. I understand that the resolution is obtained from Relion maps

@CrollTristan Those low countour "features" (why not calling it noise?) are a consequence of chunking the map into cubes, and their intensities should be negligible compared to the rest of the volume. What is the contour level?

Can any #cryoEM folk tell me what on earth is going on in the low contours of this map (EMD-32679)?

@CryoTUD @SjorsScheres @biochem_fan LocScale maps are normalized before training in DeepEMhancer. This is a standard approach in Deep Learning

@SjorsScheres @biochem_fan Possibly there has been some additional masking and histogram shaping prior to training; and remnants from striding may also be culprit - maybe @1avier knows more?

@SjorsScheres @biochem_fan DeepEMhancer maps are not supposed to be used to evaluate resolution values. I think this is explicitly said in the DeepEMhancer paper. Agree that validation is key in cryoEM

@Yoshi__Ichikawa Contrats & good luck

@Elena_Salamanca Políticos tenemos muchos, Periodistas parece que no tantos. Usted debería ser (o al menos parecer) imparcial.

Este es el tipo de gente a la que moviliza Pablo Iglesias cuando señala a un "asalariado", un trabajador/a, de una empresa que a él no le gusta. Más de 2.300 tuits en 11 horas.