Enrichment Probe Sets Combining Universal and Lineage-Specific Targets Help Resolve Recalcitrant Lineages

To resolve the phylogenetic relationships within the recalcitrant clusioid clade of Malpighiales, researchers developed the Clusioids626 probe set by combining universal Angiosperms353 targets with 273 lineage-specific nuclear orthologs, which successfully outperformed the universal panel alone in generating high-resolution nuclear and plastid data while revealing significant cyto-nuclear conflicts.

The Gist

Imagine you are trying to solve a massive, ancient family reunion photo album. You have thousands of relatives (plant species), but the photos are faded, torn, and some people look so much alike that it's impossible to tell who is related to whom. This is the challenge scientists face when studying a group of flowering plants called the Clusioid clade (which includes mangosteens, St. John's Wort, and riverweeds).

For years, trying to figure out their family tree has been a nightmare. The relationships are "recalcitrant"—a fancy word meaning they stubbornly refuse to be solved.

Here is how the authors of this paper cracked the case, explained in simple terms:

1. The Problem: The "One-Size-Fits-All" Sweater Didn't Fit

Scientists have a standard tool called Hyb-Seq. Think of this like a giant fishing net designed to catch specific DNA "fish" from any plant in the world. The most famous net is called Angiosperms353. It works great for catching a broad variety of fish, but when you try to catch very specific, tricky fish (like the Clusioid plants), the net is too loose. It misses the small details needed to tell the cousins apart.

2. The Solution: A Custom "Super-Net"

The researchers decided to build a better net. They created a new tool called the Clusioids626 kit.

• The Universal Part: They kept the original 353 "universal" targets from the Angiosperms353 net. This ensures they can still compare these plants to any other flowering plant in the world.
• The Custom Part: They added 273 new targets specifically designed just for the Clusioid family.

The Analogy: Imagine the Angiosperms353 kit is a generic t-shirt that fits everyone loosely. The Clusioids626 kit is that same t-shirt, but with custom tailoring added to the sleeves and collar to fit the Clusioid family perfectly. It's a "hybrid" tool that gives you the best of both worlds: broad applicability and high precision.

3. The Experiment: Fishing with the New Net

The team tested this new kit on 70 different plant samples, ranging from fresh leaves to dried-up museum specimens (herbarium samples).

• The Result: The new net caught way more fish than the old one.
  • The old net (Angiosperms353) caught about 198 genetic markers on average.
  • The new net (Clusioids626) caught about 481 markers on average.
  • It also caught more "information" (genetic clues) to help solve the puzzle.

4. The Big Discovery: A Family Feud (Cyto-Nuclear Conflict)

Once they had all the DNA data, they built the family tree. But here is where it gets weird.

They looked at the family tree from two different angles:

1. The Nuclear Tree: Looking at the DNA inside the plant's "nucleus" (the main control center).
2. The Plastid Tree: Looking at the DNA inside the "chloroplasts" (the parts that make food via photosynthesis).

The Twist: The two trees told different stories.

• The Nuclear DNA said: "Family A and Family B are sisters."
• The Plastid DNA said: "No, Family A is actually related to Family C!"

What does this mean?
This is like finding out that a child looks exactly like their mother (nuclear DNA) but has the exact same eye color as their father's cousin (plastid DNA). In plants, this usually happens because of hybridization (ancient mixing of species) or incomplete lineage sorting (when a family splits so fast that the genetic "hand-me-downs" get mixed up).

The paper found that the Clusioid family tree is a tangled mess of ancient mixing events, which is why it was so hard to solve before.

5. Why This Matters

• Better Tools: This study proves that combining a "universal" kit with "custom" targets is the future of plant science. It solves problems that generic tools can't.
• Solving the Mystery: They finally mapped out the relationships between the five major families in this group with high confidence.
• Understanding Evolution: The conflict between the two DNA types tells us that these plants have a wild evolutionary history involving rapid explosions of new species and ancient hybridization.

In a Nutshell

The scientists took a generic tool, upgraded it with custom parts specifically for a difficult group of plants, and used it to finally untangle a 2,000-year-old family feud. They discovered that these plants are so closely related and mixed up that their DNA tells two different stories, revealing a complex history of rapid evolution and ancient mixing.

Technical Summary

1. Problem Statement

The order Malpighiales is one of the most phylogenetically challenging (recalcitrant) groups of flowering plants due to a rapid Cretaceous radiation, resulting in a "perfect storm" of incomplete lineage sorting (ILS), introgression, and gene tree estimation errors. Within this order, the clusioid clade (~2,200 species, 5 families: Bonnetiaceae, Calophyllaceae, Clusiaceae, Hypericaceae, and Podostemaceae) has historically resisted resolution.

• Limitations of Universal Kits: The widely used Angiosperms353 universal probe set provides broad applicability but often lacks sufficient resolution for specific, rapidly radiating lineages like the clusioids.
• Limitations of Custom Kits: While lineage-specific kits offer higher resolution, they lack the universality to integrate with existing broad-scale datasets.
• The Gap: There is a need for a hybrid approach that combines the broad applicability of universal probes with the high resolution of lineage-specific probes to resolve deep and shallow nodes in difficult clades.

2. Methodology

The authors developed a novel workflow to design, validate, and apply a custom hybrid probe set.

A. Probe Design (Clusioids626 Kit)

• Target Selection: Using 12 transcriptomes (11 from the clusioid clade, 1 outgroup), the authors employed MarkerMiner and HybPiper to identify nuclear orthologs.
• Hybrid Composition: The final kit targets 626 nuclear orthologs (~6.6 Mb):
  • 217 targets unique to the clusioid clade (not in Angiosperms353).
  • 203 targets shared with Angiosperms353 but optimized for the clade.
  • 150 targets from Angiosperms353 not recovered by initial mining but included for compatibility.
• Filtering Criteria: Targets were selected based on presence in ≥7 species, inclusion of Hypericum (the largest genus), minimum length of 350 bp, and a phylo-informative site proportion (PPIC) ≥ 0.08.
• Probe Construction: 39,936 biotinylated RNA probes (120-mers) were designed with 1x tiling density.
• Plastid Recovery: A custom amino acid-based plastid target file (105 coding DNA sequences) was created to mine off-target reads for organellar data.

B. Experimental Validation

• Taxon Sampling: 70 accessions representing all five clusioid families and related outgroups (Ochnaceae, Linaceae, Celastrales).
• Sequencing: Hybridization capture (Hyb-Seq) was performed on 38 herbarium samples; 32 additional samples were mined from public databases.
• Bioinformatics: Reads were processed using HybPiper for target recovery, MAFFT for alignment, and trimAl for trimming. Paralogs were filtered using paralog_retriever.

C. Phylogenomic Analysis

• Nuclear Trees: Inferred using both Multispecies Coalescent (MSC) via ASTRAL-III and Maximum Likelihood (MLE) via IQ-TREE2 on concatenated data.
• Plastid Trees: Inferred via MLE from the 105 recovered plastid CDS.
• Conflict Analysis: Gene tree discordance was assessed using Quartet Scores (QS) and tanglegrams. Reticulation (hybridization/ILS) was explored using SplitsTree (Neighbor-Net) and PhyloNet.
• Comparison: Performance was directly compared against data processed solely with the Angiosperms353 target file.

3. Key Contributions

1. Development of Clusioids626: The first probe set designed at the supra-familial level (covering an entire clade of 5 families) that successfully integrates universal (Angiosperms353) and lineage-specific targets.
2. Optimized Workflow: A robust pipeline for designing hybrid probe sets that balances copy-number, length evenness, and phylo-informativeness to maximize recovery in degraded herbarium DNA.
3. Resolution of a Recalcitrant Clade: Provided the first fully resolved nuclear phylogeny for the clusioid clade backbone, overcoming previous topological instability.
4. Detection of Cyto-Nuclear Discordance: Systematically identified and visualized strong conflicts between nuclear and plastid evolutionary histories within the clade.

4. Key Results

A. Probe Performance

• Recovery: The Clusioids626 kit recovered a median of ~600 orthologs per sample (97% of targets).
• Efficiency: Average on-target read percentage was 50.4% (ranging from 0.2% to 86.1%). This is significantly higher than the Angiosperms353 panel alone (median 5.35% on-target reads for the same samples).
• Data Quality: The Clusioids626 dataset yielded longer alignments (mean 2,135 bp vs. 956 bp for Angiosperms353) and higher Parsimony-Informative Characters (PICs), leading to more robust topological support.

B. Phylogenetic Resolution

• Nuclear Topology: Both MSC and MLE analyses strongly supported the monophyly of the clusioid clade and its five families.
  • Topology: (Clusiaceae, (Bonnetiaceae, (Calophyllaceae, (Hypericaceae, Podostemaceae)))).
  • Support: 45 out of 66 nodes had maximum support (Local Posterior Probability = 1.0).
• Comparison: The Clusioids626 kit resolved more nodes with high support and reduced gene tree discordance compared to the Angiosperms353 panel alone.

C. Cyto-Nuclear Conflict

• Major Discordance: A strong conflict was detected regarding the placement of Bonnetiaceae and Clusiaceae.
  • Nuclear: Clusiaceae is sister to all other clusioids.
  • Plastid: Clusiaceae is sister to Bonnetiaceae.
• Internal Conflict: Significant discordance was observed within Calophyllaceae (6 of 11 genera shifted positions) and Hypericaceae (tribal relationships).
• Reticulation: SplitsTree and PhyloNet analyses revealed widespread parallel edges and reticulation, suggesting historical hybridization and ILS, particularly in Podostemaceae (aquatic) and Calophyllaceae.

5. Significance

• Methodological Advancement: This study demonstrates that hybrid probe sets (universal + lineage-specific) are superior to universal kits alone for resolving difficult, rapid radiations. It provides a scalable model for other recalcitrant clades where universal kits fail to provide sufficient resolution.
• Biological Insight: The study resolves long-standing phylogenetic uncertainties in the clusioid clade, confirming the monophyly of the group and clarifying family-level relationships.
• Evolutionary Complexity: The detection of strong cyto-nuclear conflict highlights that the evolutionary history of Malpighiales is not strictly tree-like but involves significant reticulation (hybridization/introgression) and incomplete lineage sorting, necessitating integrative genomic approaches.
• Resource Utility: The Clusioids626 kit enables the use of degraded herbarium DNA for high-resolution phylogenomics, facilitating the integration of natural history collections into modern genomic studies.