Probabilistic inference of lateral gene transfer events

Mehmood Alam Khan, Owais Mahmudi, Ikram Ullah, Lars Arvestad, Jens Lagergren

Abstract

Microbial evolution is characterized by frequent Lateral Gene Transfer (LGT), a process that complicates the reconstruction of the Tree of Life. Traditional parsimony-based methods often struggle to distinguish LGT from gene duplication and loss (GDL) or to accurately place transfer events on a species tree. We describe a robust Bayesian framework that reconciles gene trees with dated species trees using the DLTRS model (Duplication, Loss, Transfer, Rate, and Sequence evolution). By sampling from a posterior distribution, our method provides statistical confidence for inferred LGT events and identifies "LGT highways"—lineages with high rates of genetic exchange. Validation on synthetic and biological datasets demonstrates high accuracy and the ability to uncover ecologically significant genetic swaps.

Introduction: The Microbial Swap Meet

In classical evolutionary theory, genetic inheritance is primarily vertical. However, bacteria and archaea frequently engage in Lateral Gene Transfer (LGT), swapping genetic material across lineage boundaries. This mechanism is a primary driver of metabolic innovation and the spread of antibiotic resistance. For computational biologists, the challenge lies in distinguishing LGT from other processes like gene duplication and loss (GDL). Existing methods often rely on parsimony, seeking the minimum number of events to explain a phylogeny. However, biological reality is often more complex, and parsimony fails to provide statistical confidence or account for temporal constraints—ensuring a gene isn't transferred to a lineage that hasn't yet evolved.

Methodology: The DLTRS Framework

The proposed method utilizes a probabilistic approach based on the DLTRS model. Rather than simply reconstructing a gene tree, the focus is on reconciliation: mapping the nodes of a gene tree onto the branches of a dated species tree.

Technical Implementation

  • Bayesian MCMC Framework: The tool employs Markov Chain Monte Carlo (MCMC) sampling to explore the space of potential reconciliations, yielding Maximum A Posteriori (MAP) probabilities for specific events.
  • Efficiency via Rao-Blackwellisation: To manage the computational complexity of MCMC, the authors use Rao-Blackwellisation to focus sampling on the most probable realizations.
  • Relaxed Molecular Clock: A Gamma-distributed substitution rate model accounts for varying evolutionary rates across different lineages.
  • Discretization: The continuous-time species tree is discretized, enabling the use of dynamic programming to efficiently compute reconciliation probabilities.

Key Results

The framework was rigorously tested using synthetic data and two significant biological datasets: Mollicutes and Cyanobacteria.

Simulation Accuracy

In synthetic tests involving 500 gene families, the method demonstrated remarkable precision. In 129 out of 303 cases, it identified the exact number of LGT events with over 80% probability. Furthermore, it correctly pinpointed both donor and recipient lineages in a majority of tested instances.

Distribution of distances from true LGT events across simulated gene families (panels a-d represent different simulation parameters).

Biological Insights

  • Mollicutes: The analysis identified a prominent LGT highway between Mesoplasma florum and the ancestor of the Mycoplasma capricolum/mycoides group, supported by 116 gene families.
  • Cyanobacteria: The method confirmed known highways between freshwater and filamentous clades and identified new, recent exchange routes within the marine Synechococcus subclade.
  • Temporal Precision: By using a dated species tree as a scaffold, the model provided strong temporal estimates for when these genetic swaps occurred.

Discussion and Future Directions

One of the most significant advantages of this probabilistic approach is the identification of "LGT highways." These frequent exchange routes often reflect shared ecological niches, such as species co-habiting in a specific environment.

However, limitations remain. The current DLTRS model does not explicitly account for Incomplete Lineage Sorting (ILS), where ancestral variation persists through speciation. Additionally, the Bayesian MCMC approach is computationally intensive, posing challenges for scaling to massive genomic datasets. Finally, the requirement for a well-supported, dated species tree limits the method's application to less-studied microbial groups. Future work will focus on integrating ILS into the model and improving computational efficiency to handle larger-scale phylogenomic analyses.

Summary updated

Abstract

Background Lateral gene transfer (LGT) is an evolutionary process that has an important role in biology. It challenges the traditional binary tree-like evolution of species and is attracting increasing attention of the molecular biologists due to its involvement in antibiotic resistance. A number of attempts have been made to model LGT in the presence of gene duplication and loss, but reliably placing LGT events in the species tree has remained a challenge. Results In this paper, we propose probabilistic methods that samples reconciliations of the gene tree with a dated species tree and computes maximum a posteriori probabilities. The MCMC-based method uses the probabilistic model DLTRS, that integrates LGT, gene duplication, gene loss, and sequence evolution under a relaxed molecular clock for substitution rates. We can estimate posterior distributions on gene trees and, in contrast to previous work, the actual placement of potential LGT, which can be used to, e.g., identify “highways” of LGT. Conclusions Based on a simulation study, we conclude that the method is able to infer the true LGT events on gene tree and reconcile it to the correct edges on the species tree in most cases. Applied to two biological datasets, containing gene families from Cyanobacteria and Molicutes, we find potential LGTs highways that corroborate other studies as well as previously undetected examples.
BibTeX
@article{khan2016probabilistic,
  author = {Mehmood Alam Khan and Owais Mahmudi and Ikram Ullah and Lars Arvestad and Jens Lagergren},
  title = {{Probabilistic inference of lateral gene transfer events}},
  journal = {BMC Bioinformatics},
  year = {2016},
  doi = {10.1186/s12859-016-1268-2},
}
Figure