Verticall view - rrwick/Verticall GitHub Wiki

The verticall view command is an optional step to visualise some of Verticall's inner workings using interactive plots. It operates on a single assembly pair. This step is not required for any workflow, but it can help you to understand exactly why Verticall behaved the way it did for a particular pair of assemblies. When run, verticall view produces four plots, examples of which are shown below.

Note that if you are running Verticall on a remote server (e.g. connecting via ssh), then verticall view might not be able to display the plots (read more here). The easiest option is to copy the assemblies to your computer and run verticall view locally.

Example usage

Verticall view must be given a directory of assemblies and the names of which two assemblies will be compared.

verticall view -i assemblies --names a,b

Figure 1

Figure 1 shows the distance distribution coloured by the thresholds: vertical (t_low ≤ d ≤ t_high) is blue, ambiguous (t_v-low ≤ d < t_low or t_high < d ≤ t_v-high) is grey and horizontal (d < t_v-low or t_v-high < d) is red. It also shows the smoothed distribution (solid line), the mean of the full distribution (dotted vertical line) and median of the full distribution (dashed vertical line). See the Pairwise assembly comparison page for more details.

Verticall view figure 1

Figure 2

Figure 2 shows the same distribution but after ambiguous regions have been resolved to either vertical or horizontal. The mean (dotted line) and median (dashed line) in this figure are for the vertical-only component of the distribution.

Verticall view figure 2

Figure 3

Figure 3 shows the painted alignments. Solid vertical lines indicate the boundary between alignments, and the background colour indicated the classification (blue for vertical, red for horizontal).

Verticall view figure 3

Figure 4

Figure 3 shows the painted contigs. It is similar to Figure 3, except each section (separated by solid vertical lines) represents a contig in the first of the two assemblies. There is also a third possible classification, unaligned (white), for where the two genomes did not align.

Verticall view figure 4

Full help output

usage: verticall view -i IN_DIR -n NAMES [--window_count WINDOW_COUNT] [--window_size WINDOW_SIZE]
                      [--ignore_indels] [--smoothing_factor SMOOTHING_FACTOR] [--secondary SECONDARY]
                      [--verbose] [--index_options INDEX_OPTIONS] [--align_options ALIGN_OPTIONS]
                      [--allowed_overlap ALLOWED_OVERLAP] [--sqrt_distance] [--sqrt_mass]
                      [--result RESULT] [--vertical_colour VERTICAL_COLOUR]
                      [--horizontal_colour HORIZONTAL_COLOUR] [--ambiguous_colour AMBIGUOUS_COLOUR]
                      [-h] [--version]

view plots for a single assembly pair

Required arguments:
  -i IN_DIR, --in_dir IN_DIR       Directory containing assemblies in FASTA format
  -n NAMES, --names NAMES          Two sample names (comma-delimited) to be viewed

Settings:
  --window_count WINDOW_COUNT      Aim to have at least this many comparison windows between assemblies
                                   (default: 50000)
  --window_size WINDOW_SIZE        Use this defined window size for all pairwise comparisons (default:
                                   dynamically choose window size for each pair)
  --ignore_indels                  Only use mismatches to determine distance (default: use both
                                   mismatches and gap-compressed indels)
  --smoothing_factor SMOOTHING_FACTOR
                                   Degree to which the distance distribution is smoothed (default: 0.8)
  --secondary SECONDARY            Peaks with a mass of at least this fraction of the most massive peak
                                   will be used to produce secondary distances (default: 0.7)
  --verbose                        Output more detail to stderr for debugging (default: only output
                                   basic information)

Alignment:
  --index_options INDEX_OPTIONS    Minimap2 options for assembly indexing (default: -k15 -w10)
  --align_options ALIGN_OPTIONS    Minimap2 options for assembly-to-assembly alignment (default: -x
                                   asm20)
  --allowed_overlap ALLOWED_OVERLAP
                                   Allow this much overlap between alignments (default: 100)

Plot settings:
  --sqrt_distance                  Use a square-root transform on the genomic distance axis (default:
                                   no distance axis transform)
  --sqrt_mass                      Use a square-root transform on the probability mass axis (default:
                                   no mass axis transform)
  --result RESULT                  Number of result to plot (used when there are multiple possible
                                   results for the pair, default: 1)

Colours:
  --vertical_colour VERTICAL_COLOUR
                                   Hex colour for vertical inheritance (default: #4859a0)
  --horizontal_colour HORIZONTAL_COLOUR
                                   Hex colour for horizontal inheritance (default: #c47e7e)
  --ambiguous_colour AMBIGUOUS_COLOUR
                                   Hex colour for ambiguous inheritance (default: #c9c9c9)

Other:
  -h, --help                       Show this help message and exit
  --version                        Show program's version number and exit