• MycoCosm/PhycoCosm Portal Help
• Species Tree

Tree-building Methods

1. MCL http://micans.org/mcl
2. MAFFT http://mafft.cbrc.jp/alignment/software
3. Timal http://trimal.cgenomics.org
4. FastTree http://www.microbesonline.org/fasttree

Tree

Tree view represented by two major parts: tree and table. Tree side represents annotated species tree created using the tools described above. Representation of the tree side is standard. Distance to parent and support value both represented here. Distance shown as a length of the link to the parent node while support as color label based on scale. Support value scale has following labels:

• exact 1.0
• value between 0.7 and 1.0
• value below 0.7
• no support value; please note that this is not the same like 'zero' support. For example when we stitches together two part of the tree, or create artificial root node at the point where it need to be, we do not have a support value for this split.

Exact values of distance to the parent and support could be viewed on tree node selection panel when given node is selected.

Table (right side of the screen) shows list of individual species in tabular form. Columns shows

• Species/strain name
• Assembly Length (in bp)
• Total number of genes in filtered set
• Reference to the publication, if paper was published

Tree Controls

Each node on the tree could be highlighted by hover over the node with mouse. Following operations are available on any highlighted node:

• simple click with mouse button - toggle node (expand/collapse); last expanded/collapsed node become selected
• Ctrl+mouse click - select or de-select node without expand or collapse it
• Shift+mouse click - "zoom into" the given node; node become the "virtual root" of the view

Selected Tree Node Panel

Selected node panel always appears on top right corned of the browser window. It appears when any node is selected. It has following information:

• Node Id; it could be artificial id, like "n335", or species/strain portal id, like "Penla1"
• Label; some node are annotated with labels. Those are ether leaf nodes – in this case label will be species/strain name for given portal. Or it could be some NCBI taxonomy names, that we assigned to the given node (see NCBI Taxonomy Annotations section below)
• Support; this is support value for given split, computed by the tree computation process.
• Distance; this is distance to parent node, computed by the tree computation process.
• Total Leaf; total number of leaf (species/strain) nodes below selected node.
• NCBI Taxon; assigned NCBI taxon on given node, see NCBI Taxonomy Annotations section below.
• NCBI Alias; assigned NCBI taxon alias on given node, see NCBI Taxonomy Annotations section below.
• Taxon Support; NCBI taxonomy support value, see NCBI Taxonomy Annotations section below.

Selected node panel controls

• move current tree "root" one level up
• make currently selected node a new "root" for the tree
• clear selection and hide selected node panel

NCBI Taxonomy Annotations

We annotate our species tree with NCBI taxonomy labels using "best bidirectional match with cutoff" approach. For each node we compute list of it leaf nodes (with known NCBI taxonomy) and than look on NCBI taxonomy tree for the node that has most of those nodes. It could be that we have "perfect" match - all of leaf nodes belong to one or more NCBI nodes. Or in some cases it is "less than perfect"" match, but still the best across all possible pair of ours and NCBI nodes and ratio of "matched" leaf nodes to "total" leaf nodes is grater than a cutoff value. This ratio become "Taxon Support" value. The top-most NCBI node become an assigned NCBI taxon for given node on our tree. If we have more than one best matches from the same hierarchy - all lower taxons become Taxon Aliases, we basically cannot say on what level of NCBI taxonomy between all aliases our node must be placed.

We do our best to assign valid and meaningful NCBI labels to nodes on our tree, however, as expected, majority of nodes on our tree does not have NCBI labels.

Top Controls

On the top of the tree view there are several important controls.

To Default

This button will restore 'default' expand/collapse state of the tree. Usually that mean some nodes are expanded and some are collapsed to give you the 'best' visual representation of the tree.

Expand All

This button will expand all nodes on the tree. Some trees could be big.

Filter/Clear

This control let you search across nodes. You can enter some text and nodes will be searched across all tree for the match. The text could be, for example, species name, or NCBI taxon. Filtered result shows all nodes that match selection and hide or 'deem' all other nodes. Clear button removes filtering, but does not change expand/collapse state of the tree.

Ladderize

Use of this control will cycle over 4 possible modes of tree 'ladderization'. This will change the order of species/strand rows on table view. Ladderization mean soring children nodes on each tree node by certain criteria. 2 modes related to how we sort nodes: (1) by total distance to the root; (2) by sum of all distances to the root for nodes below given. Results will vary depending of specific tree. This two modes may be ascending ore descending giving total 4 possible options.

To Root

This button will move current 'virtual' tree root to the 'actual'tree root. Usually that mean you will see whole Fungi tree.

Reload

This button will reload tree data structure from the underlying storage.