Create your own conference schedule! Click here for full instructions

The Virtual Conference is located at https://botany2021.pathable.co/.

Abstract Detail



Mycology

Lin, Pei-Yu [1], Huang, Yu-Chun [2], Chen, Pao-Yang [3].

500 Fungal Genomes: Finding Fungal Essential Genes based on Core Genes Prediction.

The fungal core genes (CGs) presenting in multiple fungal genomes are predictive of the essential genes (EGs) that are critical for growth and viability. EGs are often used for identifying potential antifungal drug targets, constructing phylogenomic tree for precise taxonomy, as well as predicting biosynthetic gene clusters (BGCs) through the resistance-gene-based hypotheses. There was still no CG database for fungi, and the current fungal EG databases were built from only three fungal species based on labour-intense experiments. Here we constructed a fungi-specific CG database (FunCore), containing 491 fungal genomes, across 68% of all assembled genomes, kindly shared by communities, making FunCore the largest CG database. We implemented a network-based clustering method, Louvain algorithm, effectively cluster 6M proteins by sequence similarity. The clustering resulted in 95,259 gene subnets; each represents a group of similar proteins across several species. We exploited each gene subnet as a network and used their network structures as parameters to build a generalized linear model that altogether allows accurate ranking of subnets by their probability of being an EG. The predicted probability scores are further used for building up the phylogenomic tree in association with the fungal taxonomy. We found that there were two types of EGs, the global EGs exhibiting in most species and the local EGs only vital for a few fungi sharing the same clade on phylogenomic tree, suggesting that these local EGs are unique to specific fungal families for living. As a validation, our methods reached 99% accuracy specifically in predicting global EGs. Our top predictions of EG subnets identified novel CG subnets cover numbers of species, and are associated with transport functions which are the characters for EGs. Additionally, a yeast EG KRE33 known as a toxin resistance gene was found in a high-rank subnet that can be directly used in BGCs prediction. FunCore is able to take query protein sequences, and returns the likelihood of the query being an EG and their biological functions. It also displays a dynamic 3D network graph for users to visualize the homologus proteins across multiple species within a CG subnet. The mouse-over function of FunCore pops up photos of fungi, so to connect the morphologic features among similar proteins as inferences for taxonomy. FunCore covers the largest number of fungal genomes, using a novel network biology prediction and statistical modelling, giving a rank of fungal CGs as a valuable resource for fungal genomic research.


Log in to add this item to your schedule

1 - Academia Sinica, Institute of Plant and Microbial Biology, 128 Academia Road, Section 2, Nankan Dist., Taipei, 11529, Taiwan
2 - Academia Sinica, Institute of Plant and Microbial Biology, 128 Academia Road, Section 2, Nankan Dist., Taipei, 11529, Taiwan
3 - Academia Sinica, nstitute of Plant and Microbial Biology, 128 Academia Road, Section 2, Nankan Dist., Taipei, 11529, Taiwan

Keywords:
core genes
essential genes
fungi
database
Mycology.

Presentation Type: Poster
Session: MYP1, Mycology Posters I
Location: /
Date: Monday, July 19th, 2021
Time: 5:00 PM(EDT)
Number: MYP1001
Abstract ID:99
Candidate for Awards:None


Copyright © 2000-2021, Botanical Society of America. All rights reserved

aws4