Khanh Dao Duc

CAT PETR: A Graphical User Interface for Differential Analysis of ProteinPhosphorylation Data biorxiv.org/content/10.110…

RiboXYZ: a comprehensive database for visualizing and analyzing ribosome structures academic.oup.com/nar/advance-ar…

RiboXYZ: A comprehensive database for visualizing and analyzing ribosomestructures biorxiv.org/cgi/content/sh… #bioRxiv

RiboXYZ: A comprehensive database for visualizing and analyzing ribosome structures biorxiv.org/content/10.110…

1/ Excited to announce our lab’s first paper at @ubclifesciences! Loss of the RNA binding protein FMRP is the most common heritable cause of autism. We find that FMRP has one major function that is conserved from fly eggs to mammalian brains. bit.ly/3PxHGeS

Viral surface geometry shapes influenza and coronavirus spike evolution biorxiv.org/cgi/content/sh… #bioRxiv

Welcoming feedback on using MorphOT to generate movies between your cryoEM maps! Plugin written for the great @UCSFChimeraX

MorphOT: Transport-based interpolation between EM maps with UCSF ChimeraX biorxiv.org/cgi/content/sh…

Our article on comparative analysis of the ribosome across 20 different species just got published: academic.oup.com/nar/advance-ar…

In this paper we identify the key parameters that govern translation efficiency, and formulate four design principles to optimize it in terms of protein production rate and resource usage. biorxiv.org/content/early/…

Outstanding work by my postdoc @kdaoduc on a comparative analysis of the ribosome exit tunnel geometry across 20 species. Has implications on the evolution of translation dynamics and inter-species differences in several co-translational processes. biorxiv.org/content/early/…

Efficiently inferring the demographic history of many populations with allele count data biorxiv.org/content/early/…

@jgschraiber @yun_s_song In practice, this measure of trans_eff may be biased for high density genes, but we actually proposed a way to correct it in this paper :) : journals.aps.org/pre/abstract/1…

@jgschraiber @yun_s_song In theory trans_Eff_1 < trans_eff_2 implies that J_1<J_2, but the relation is not linear especially when the density increases. What we can do in addition is to exactly compute J from trans_eff.

Super excited about this work. Closed-from formulas for steady-state local ribosome densities, currents (overall protein production rate), and phase transition boundaries for a general model of mRNA translation with inhomogeneous elongation rates. arxiv.org/abs/1803.05609

@jgschraiber @yun_s_song I would say that if the site you pick is associated with high speed, the observed density could in practice be more sensitive to noise or affected by the sample size so it's better to pick a site with relatively low lambda_k.

@kdaoduc @yun_s_song Ahhh I think I see. Fixing one rate lets use equation 2 to get J. That makes sense. Is there any benefit to lambda_k = 1 vs. fixing J = 1?

@jgschraiber @yun_s_song Should say about the second message I sent that the depletion measuremment actually does not directly give J but rather the time scale

@jgschraiber @yun_s_song Yes, this is more or less what we do except we fix one local rate lambda_k to be 1 and get all the other rates and flux relative to this one

@jgschraiber @yun_s_song In practice, for ribosome profiling data, you can use some measurement of ribosome depletion to get J (like in fig 3b of Ingolia et al. Cell 2011). Which also means our method helps to get the rates 2/2

@jgschraiber @yun_s_song J is the effective flux of particles i.e. how many particles cross a site per unit of time. As the model is invariant by time scaling (scaling all rates by K gives a flux K*J), you need to fix a time scale if you want an absolute measure of it. 1/2