Graphs

Platt's graph algorithm:

Before we build our own graphs, let's use a basic graph example

state_graph = data.frame(from = c("early notochord progenitor", 
                                  "early notochord", 
                                  "early vacuolated notochord",
                                  "early notochord", 
                                  "early notochord sheath"), 
                         to = c("early notochord", 
                                "early vacuolated notochord", 
                                "late vacuolated notochord", 
                                "early notochord sheath", 
                                "late notochord sheath")) %>% 
                         igraph::graph_from_data_frame() 

plot(my_graph)

Making a cell_state_graph object

The new_cell_state_graph function takes the following as input:

• cell_state_graph - an igraph object
• ccs - a Hooke cell_count_set object

notochord_state_graph = new_cell_state_graph(state_graph, ccs)
plot_annotations(notochord_state_graph, node_size = 4.5)

Larger graphs

if you have a larger graph that you want to plot by tissue level annotation, you can specify a grouping in new_cell_state_graph and the number of layers you want to arrange each group in

ref_state_graph = new_cell_state_graph(combined_state_graph, 
                                       ref_ccs, 
                                       group_nodes_by="projection_group", 
                                       num_layers=3)


plot_by_annotations(ref_state_graph) + theme(legend.position = "none")

Manipulating graphs

Get the parents:

• cell_state_graph
• cell_state

get_parents(cell_state_graph@graph, cell_state)

For example:

> get_parents(notochord_state_graph@graph, "early vacuolated notochord")

returns

> "early notochord"

Get the children:

• cell_state_graph
• cell_state

get_children(cell_state_graph@graph, cell_state)

For example:

> get_children(notochord_state_graph@graph, "early vacuolated notochord")

returns:

> "late vacuolated notochord"

Get the siblings:

• cell_state_graph
• cell_state

get_siblings(cell_state_graph@graph, cell_state)

For example:

> get_siblings(notochord_state_graph@graph, "early vacuolated notochord")

returns

> "early notochord sheath"

Constructing a graph

We first construct a graph on clusters...

cds = cluster_cells(cds, random_seed = 42, res=1e-3)
colData(cds)$cell_state = monocle3:::clusters(cds)

cluster_res = assemble_partition(cds, 
                                   partition_name = "pectoral fin",
                                   sample_group = "embryo",
                                   cell_group = "cell_state",
                                   interval_col = "timepoint",
                                   component_col = "assembly_group",
                                   perturbation_col = "perturbation",
                                   ctrl_ids = c("ctrl-inj"),
                                   num_threads = 6,
                                   batch_col = "expt")

...then contract the graph, and use it as a prior for a graph built on cell types...

contract_graph = platt::contract_state_graph(ccs, cluster_res$mt_graph[[1]], group_nodes_by = "cell_type")

global_wt_graph_edge_allowlist = igraph::as_data_frame(contract_graph)
global_wt_graph_edge_allowlist = global_wt_graph_edge_allowlist %>% select(from, to) %>% distinct()


cell_type_res = assemble_partition(cds, 
                                   partition_name = "pectoral fin",
                                   sample_group = "embryo",
                                   cell_group = "cell_type",
                                   interval_col = "timepoint",
                                   component_col = "assembly_group",
                                   perturbation_col = "perturbation",
                                   edge_allowlist = global_wt_graph_edge_allowlist, 
                                   ctrl_ids = c("ctrl-inj"),
                                   num_threads = 6,
                                   batch_col = "expt")

The output is as follows:

data	wt_graph	mt_graph	perturbation_effects	perturbation_table
<tibble [29,181 × 6]>	<igraph>	<igraph>	<tibble [27 × 2]>	<tibble [1,422 × 23]>

You can extract the graph from the dataframe to construct a cell_state_graph.

pf_graph = partition_res$mt_graph[[1]]
pf_ccs = new_cell_count_set(pf_cds, cell_group = "cell_type", sample_group = "embryo")
pf_csg = new_cell_state_graph(pf_graph, pf_ccs)

plot_annotations(pf_csg)

For more information about plotting on a Platt graph, see our Plotting page