Reconstruct tree by SCITE

This example shows how to construct a phylogenetic tree using SCITE on a binary single-cell genotype matrix.

import scphylo as scp

# sphinx_gallery_thumbnail_path = "_static/thumbnails/scite.png"

First, we load a binary test single-cell genotype data.

df_in = scp.datasets.test()
df_in.head()

	mut0	mut1	mut2	mut3	mut4	mut5	mut6	mut7	mut8	mut9	mut10	mut11	mut12	mut13	mut14	mut15	mut16	mut17	mut18	mut19
cellIDxmutID
cell0	1	1	1	1	1	1	1	0	3	3	0	3	0	0	0	3	0	0	0	0
cell1	3	0	0	0	0	0	3	0	1	1	3	1	0	0	3	0	0	1	1	1
cell2	0	0	0	0	0	0	0	1	1	3	1	1	0	0	3	1	1	1	1	0
cell3	0	0	0	0	0	0	0	3	1	0	0	1	0	0	0	1	1	1	0	0
cell4	0	0	0	0	0	0	0	1	1	1	1	1	0	0	3	0	0	0	0	0

Next, using scphylo.tl.scite() we remove the single-cell noises from the input.

df_out = scp.tl.scite(df_in, alpha=0.0000001, beta=0.1, n_restarts=3, n_iters=1000)
df_out.head()

running SCITE with alpha=1e-07, beta=0.1, n_iters=1000, n_restarts=3
input -- size: 20x20
input -- 0: 226#, 56.5%
input -- 1: 94#, 23.5%
input -- NA: 80#, 20.0%
input -- CF: False
output -- size: 20x20
output -- 0: 256#, 64.0%
output -- 1: 144#, 36.0%
output -- NA: 0#, 0.0%
output -- CF: True
output -- time: 0.0s (0:00:00.015396)
flips -- #0->1: 27
flips -- #1->0: 0
flips -- #NA->0: 57
flips -- #NA->1: 23
rates -- FN: 0.223
rates -- FP: 0.00000000
rates -- NA: 0.200
score -- NLL: 72.07370588267617
score: -73.64423601943437
beta: 0.4097172511966814

	mut0	mut1	mut2	mut3	mut4	mut5	mut6	mut7	mut8	mut9	mut10	mut11	mut12	mut13	mut14	mut15	mut16	mut17	mut18	mut19
cellIDxmutID
cell0	1	1	1	1	1	1	1	0	0	0	0	0	0	1	0	0	0	0	0	0
cell1	0	0	0	0	0	0	0	1	1	1	1	1	0	0	0	1	1	1	1	1
cell2	0	0	0	0	0	0	0	1	1	1	1	1	0	0	0	1	1	1	1	0
cell3	0	0	0	0	0	0	0	1	1	1	1	1	0	0	0	1	1	1	1	0
cell4	0	0	0	0	0	0	0	1	1	1	1	1	0	0	0	0	1	0	0	0

Finally, using scphylo.ul.is_conflict_free_gusfield() we check whether the inferred genotype matrix is conflict-free or not.

is_cf = scp.ul.is_conflict_free_gusfield(df_out)
print(is_cf)

True

Estimated memory usage: 354 MB