Energy costs act as weights to evolutionary paths.
Previous Project Menu:
Background
The emergence and de novo construction of proteins are elusive. The vast majority of theoretically plausible polypeptide chains fail to fold into ordered domains, let alone confer functionalities. Consequently, protein evolution from preexisting short peptides (“themes”) offers apparent advantages over random emergence. Therefore, themes are crucial entities on which we study the evolutionary origin of protein domains. These short segments have been utilized to explore the remote homology (e.g. Rossmann - P-loop; OB - SH3; IgG-like - beta-trefoil) which is not revealed in the previous protein classification system.
Ligands are required for proteins to function. The laws of interaction between proteins and ligands can indicate how ligand binding emerged and evolved. Protein evolution can be constrained or facilitated by protein-ligand interactions, so vestiges of protein origins may be found in ancient protein-ligand interactions.
Outputs: Qiu K, Ben-Tal N, Kolodny R. Similar protein segments shared between domains of different lineages, accepted in Protein Science, 2022
(1) We combined two topics "themes" and "ligand-protein interactions", selected proteins that bind ligands such as ATP, NAD, and SAM as model systems, and focused on the conservation and evolution of ligand-binding sites and ligand-binding themes of proteins. At the binding site level, we established a computational pipeline to achieve the comparison of binding modes of ligand groups (such as adenine and ribose), and found a large number of conserved structural motifs and binding modes. At the theme level, we analyzed the distribution architecture of themes in protein databases and utilized themes to understand and interpret the evolution of specific fold families (such as Rossmanns) and the emergence of ligand-binding capacity of proteins.
(2) Together with Prof. Rachel Kolodny in Haifa Univ., we complied a new bridging theme dataset with a similar procedure to a previous work of Nir, Rachel, and Danny. We discovered many interesting cases in these bridging themes and a lot of interesting stories will be uncovered in the near future!
Background
Recently, Vaishnav et al. developed a machine learning framework that can accurately predict the activity of promoters according to the input sequences. This is not only a great revolution in synthetic biology but also contributes significantly to the understanding of evolutionary biochemistry. It's the first time that we can have access to such a comprehensive promoter activity landscape via reliable prediction, and thus we can use it to explore many evolutionary questions!
Outputs: N/A at present
Together with a graduate student Siliang Song in Zhang lab, we conduct some analyses on the comprehensive promoter activity landscape.
(1) I characterized the epistasis coefficients (including high-order epistasis), γ, adaptive steps, ruggedness/slope, peaks of the promoter landscape using different mathematical models, and compared these measurements with those of null models, e.g. NK model and House-of-Card model.
(2) I conducted in silico evolution of promoter sequences and compare properties, including mutational robustness, environmental robustness, regulatory complexity between evolved promoters, random promoters and natural promoters.
Background
Microtubules (MTs) are cylindrical polymers of ab-tubulin that display pseudo-helical symmetry due to the presence of a lattice seam of heterologous lateral contacts. Cilia/flagella are composed of nine doublet microtubules surrounding a pair of singlet microtubules called the central pair (CP). Together, this arrangement forms a canonical and highly conserved 9+2 axonemal structure. The CP, which is a unique structure exclusive to motile cilia, is a pair of structurally dimorphic singlet microtubules decorated with numerous associated proteins.
2021 Mitacs Globalink Internship
Department of Anatomy and Cell Biology, McGill University, Canada, Feb. 2021 - May 2021 (Advisor: Khanh Huy Bui)Outputs: In preparation
Together with Gregor Sommer (Now a graduate student at Freie Universität Berlin), we took the responsibility for image processing and modeling of microtubules and central pairs.
(1) We fulfilled an image processing procedure of helical processing in cryoSPARC, with a special focus on 2D classification. Briefly speaking, we tested different 2D class pickers, such as filament tracer designed specifically for helical processing, regular picker, manual picker, and what we cared about most - Topaz picker, and various combinations of these methods. We finally confirmed a workflow with specific parameters in these steps.
(2) For the modeling, we were trained on the most common and popular approaches to cryo-EM modeling. I combined Alphafold2, Coot, Phenix to determine the models of two microtubule inner proteins, Rib72a/b.
Undergraduate Training
Wuhan Institute of Virology, Chinese Academy of Sciences, China, Dec. 2018 - Jan. 2021 (Advisor: Peng Gong)Outputs: Submitted to Journal of Virology
Designed and performed molecular cloning for further protein expression
Optimized the expression and purification procedure according to extant protocol and literature
Conducted complex assembly(nsp7-nsp8-nsp12-RNA and NS3-NS5) and enzymatic activity/stability assay
Crystallized elongation complex and solved structures with the help of many software
Background
Nosocomial infection and ventilator-associated pneumonia
Through investigations, we deemed that nosocomial infections, or hospital-acquired infections (HAIs), were troublesome issues in the epidemic but usually ignored by the public. According to the definition in encyclopedia, nosocomial infections refer to infections during the stay in hospital, including those caused by technical measures of surgery, therapy, diagnosis and prophylaxis, such as IDVC, injection, transfusion, inhalation, and burn treatment. The advanced diagnostic methods have revealed astonishing data on nosocomial infections, as a preprint manuscript showed the specimens collected from Renmin Hospital of Wuhan University between November 2018 and November 2019. Having consulted the researchers participating in the battles against coronavirus, we found that nosocomial infections in the epidemic were even more pervasive than usual, to a degree worsened by the attack of coronavirus, and multi-resistant gram-negative pathogen bacteria were quite considerable threats.
Ventilator was indeed a hot-spot word across the globe during the spread of this contagious pulmonary disease, for its capability to assist the patients to breathe. However, when medical staff struggles to cope with the surging positive cases, the massive use of ventilators might lead to unexpected troubles. Ventilator-Associated Pneumonia (VAP) is a severe health problem among immunocompromised patients caused by inhaling pathogen bacteria during ventilation, which subsequently contribute to high-density biofilm formation in the respiratory tract and even lead to death. Although noninvasive mechanical ventilation could to a degree prevent VAP, the mass use of invasive ventilators was inevitable, resulting from high requirements of air pressure in numerous cases. Given broad and massive applications, VAP included a large proportion of nosocomial infections in the COVID-19 pandemic, and we aimed to solve this problem.
Probiotics and human immunity
Current interests towards the effects of probiotics on human respiratory tract and their interactions with immune systems are increasing, and therapies of intaking probiotics, such as Lactobacillus spp., Bifidobacterium spp., Bacteroides spp. and Escherichia coli Nissle 1917, by nasal or oral approaches to protect the hosts from respiratory infections are proven effective in animal models. The relevant mechanisms remained largely unelucidated, while several researchers guessed that probiotics secreted chemicals to stimulate and activate the immune systems. Interestingly, probiotics are considered as next-generation chassis in biomedical fields and have been engineered to facilitate emerging therapies, on the basis of their original beneficial functions. Synlogic, the leading enterprise renowned for synthetic biotic medicines, has already promoted engineered probiotic therapies in clinical trials to address metabolic diseases, inflammations and tumors. Besides reinforcing immune systems, engineered probiotics were also harnessed to directly solve pathogen infections in animal models.
Quorum sensing and The Negotiator
It is known that many gram-negative pathogen bacteria coordinate their virulence behaviors in a cell density-dependent manner known as Quorum Sensing (QS). In this communication process, pathogen bacteria exchange autoinducers (AIs) to regulate the expression of virulence genes. When the concentration of AIs increases, the pathogen bacteria will express virulence synchronously, such as forming biofilms and secreting toxins. The interaction between pathogen bacteria and hosts is considered an arms race, with self-evolving strategies implemented respectively (e.g. biofilm formation for bacteria, and pathogen-killing immune cells recruitment for hosts). We can imagine immune cells as police squads, pathogen bacteria as criminals and autoinducers as communication tools, and in most cases the immune system of a healthy individual will eradicate the intruders. However, considering those immunocompromised patients that have suffered from the coronavirus, the balance is broken due to their impaired immune system, and criminals might easily dominate the battleground with the help of their quorum sensing systems.
Thus, we came up with an idea: how about creating negotiators (engineered probiotics) to parley with criminals (pathogen bacteria) while give instructions secretly to police squads (immune cells)? The idea later turned into an iGEM project, which our team named "The Negotiator" to salute a classic gangster movie released in 1998.
P.S. This background information was attained from our website.
2020 iGEM Competition
College of Life Sciences, Wuhan University, China, Nov. 2019 - Nov. 2020, (Advisor: Zhixiong Xie and Yu Chen)Outputs: Gold Medal
I was responsible for the mathematical modeling part of 2020 iGEM competition for the team of Wuhan University.
(1) With SimBiology in MATLAB and Mathematica, we fulfilled an ODE and a PDE framework for our quorum sensing and quorum quenching systems. To illustrate this system further, I utilized a cell programming language - gro - to explore how the spatial distribution of bacteria influences the efficiency of quorum quenching.
(2) Since we worked on enzymes whose substrates vary, I tried phylogenetic reconstruction and then ancestral protein reconstruction of these enzymes, based on the assumption that ancient enzymes may be more promiscuous. After this, I conducted molecular docking and structural analysis to analyze possible mutated positions that can empower the promiscuity of the enzymes we will use in the wet lab.
(3) I also depicted the course of fibrosis of the lung using ODEs.
For more information, please see our website.