Glossary

A phylogenetic tree is also known as an "evolutionary tree" and displays as a branching diagram, resembling a tree, diagramming relationships of genetic similarities and differences. The diagram provides a hypothesis of evolutionary history, where the tips, or "leaves", represent the end or present time.

• Node: A node represents a branching point in a tree. Each node can be connected to many children, while a binary tree constrains the number of children to exactly two descendants.

• Root: A rooted phylogenetic tree is a directed tree that starts with a unique node, the root, which corresponds to the common ancestor of all the entities at the tips or leaves of the tree, descending from this ancestor.

The branching of a phylogenetic tree depicts the branching history of common ancestry, where the pattern of branching, or topology, informs relationships. Branches can be rotated and display equivalent information as seen in the above image.

This tips for tree reading can help avoid common misinterpretations. For instance, the resource included the top image to demonstrate how branches can be rotated.

Since rotations can display equivalent information, you cannot assume that terminal taxa located more closely to one another are more closely related evolutionarily. One must follow the branches to identify the most common ancestors. For instance, in the below image, the circle and the star are less related, although adjacent in the tree on the left. Instead the star and the rectangle, although separated by the triangle, share a younger common ancestor compared to the circle, and are more related.

To identify introductions and clusters in the data, a regional index is calculated. The formula and figure shown here are from the paper introducing the cluster tracker tool. The regional index provides a weighted summary for descendents of a phylogenetic tree.

A binary model of region membership can approximate the intuitive idea that if all descendents of a node were found in region A, likely that the ancestor of that internal node was circulating in region A. The Li and Lo refer to the number of leaves (samples) that are inside or outside when looking at a focal node. The Di and Do refer to the total branch lengths to leaf descendents in the focal region, where the total branch length is equal to the mutations between the query node and a descendant leaf.

Looking at the below figure one, and the focal node, you can see that only one mutation has happened since the top sequence in blue, with only one leaf. Looking at the other branch, you can see that there were first three [3] mutations accumulated, displayed as a longer branch, before a node with a further mutation [1] leading to a total of 3 red leaves. Intuitively we can sense how the red samples are grouped together, more time has passed for all of these mutations to happen in this descendent group. The regional index calculation gives the focal node the value of 0.43, below 0.5, suggesting the blue leaf is out-of-region, with an introduction below the root, while noting the ancestor of the downstream in-region sample cluster exists along that branch as well.

Taxonium is a tool for exploring phylogenetic trees, including those with millions of nodes. Taxonium is especially powerful when applied to a tree that has been annotated with mutations. The "View Cluster" link in the CA Big Tree Cluster Tracker will open a view in Taxonium.

On the right-hand side in Taxonium, you can change the "Color by:" option from "Region" to "PANGO lineage" or "None." By default, the "Search" section will have a selected "Cluster" name from clicking the "View Cluster" link in Cluster Tracker. By clicking the small magnifying glass next to "# results" you can zoom in on the cluster. You can also click the "Add a new search" button and add on searches for other metadata such as specific accessions, other cluster names, or specific regions. The view of the tree can be zoomed in and out both vertically and horizontally with the many different magnifying glass icons on the bottom. You can also use your mouse scroll wheel to zoom in and out vertically without the icons.

With an item selected, details will appear on the right. After the name, you can click a little arrow pointing up to the left. It allows selecting the entire branch of the tree one level up. Once up a branch, a curved arrow that goes to the right should appear after a phrase such as "Number of descendants: #". Once clicked, a pop-up will offer to "List all tips" for this branch. An option to change the default selection of "name" to "meta_region" or other metadata selections will list all the regions, or selected metadata, in this branch. Options to download the JSON for the branch, or view it in NextStrain or CovSpectrum are also available by clicking this arrow.

UShER stands for Ultrafast Sample placement on Existing tRee and is the tool that puts the ever-growing new SARS-CoV-2 mutations arising in the natural environment into context on an existing phylogenetic tree. The UCSC Pathogen Genomics group uses mutation-annotated tree utilities, or matUtils, with parallel dynamic programming to quickly extract information hidden in the daily growing tree built by UShER. The phylogenetic tree updated daily by UShER is also used by the Phylogenetic Assignment of Named Global Outbreak Lineages (PANGOLIN) software to implement a dynamic nomenclature, known as the PANGO nomenclature, to classify genetic lineages for SARS-CoV-2.

UShER works with mutation-annotated trees, as shown in the figure below. The mutation-annotated tree object carries sufficient information to derive parsimony-resolved genotypes for any tip of the tree using the sequence of mutations from the root to that tip. For example, in the below figure, S5 can be inferred to contain variants G1149U, C7869U, G3179A from the first node and A2869G, specific to S5, all in relation to the reference sequence at the root. UShER’s mutation-annotated tree approach is compact and is what helps make it fast in placing new samples based on the variants they contain.