{% extends "base_QC.html" %} {% block title %}Multivariate Report ({{ item.ReportType }}) For: {{ item.Name }} {% endblock %} {% block content %}

Dataset

Original data consists of {{ item['Nsamples'] }} samples and {{ item['Nfeatures'] }} features

PCA Analysis

PCA Model Parameters

PCA QC Outputs

Figure 1. PCA scree plot of variance explained by each component (cumulative).

PCA scree plot of variance explained by each component (cumulative).

Figure 2. PCA scores plots coloured by sample type.

{% for key, value in item['QCscores'].items() %}
PCA scores plots coloured by sample type.
{% endfor %}

Figure 3. Distribution in total distance from origin (scores space) by sample type.

Distribution in total distance from origin (scores space) by sample type.
{% if 'Noutliers_strong' in item %}

Excluding samples with total distance from origin values exceeding the {{ item['scores_criticalVal'] }} limit would result in {{ item['Noutliers_strong'] }} exclusions.

{% endif %}

Figure 4. Distribution in distance from model (DmodX) by sample type.

Distribution in distance from model (DmodX) by sample type.
{% if 'Noutliers_moderate' in item %}

Excluding samples with DmodX values exceeding the {{ item['dModX_criticalVal'] }} limit would result in {{ item['Noutliers_moderate'] }} exclusions.

{% endif %} {% if 'Noutliers_total' in item %}

Excluding outliers (as specified) would result in {{ item['Noutliers_total'] }} exclusions.

{% endif %} {% if 'Outliers_total_details' in item %}

Details of potentially outlying samples

{{ item['Outliers_total_details'].to_html() }} {% endif %}

Figure 5. PCA loadings.

{% for key, value in item['loadings'].items() %}
PCA loadings.
{% endfor %}

Distribution of Values in each Metadata Field

Figure 6. Histograms of metadata distributions (plotted for fields with non-uniform values only).

{% if 'metadataDistContinuous' in item %}

Continuous data.

{% for key, value in item['metadataDistContinuous'].items() %}
Histograms of metadata distributions.
{% endfor %} {% endif %} {% if 'metadataDistCategorical' in item %}

Categorical data.

{% for key, value in item['metadataDistCategorical'].items() %}
Histograms of metadata distributions.
{% endfor %} {% endif %} {% if 'metadataDistDate' in item %}

Date data.

{% for key, value in item['metadataDistDate'].items() %}
Histograms of metadata distributions.
{% endfor %} {% endif %}

Association of PCA Scores with Metadata

Calculations Performed

Figure 7. Heatmap of correlation to PCA scores for suitable metadata fields.

{% if 'sigCorHeatmap' in item %}
Heatmap of correlation to PCA scores for suitable metadata fields.
{% else %}

{{ item['Ncorr'] }} fields where correlation to PCA scores calculated.

{% endif %}

Figure 8. Heatmap of Kruskal-Wallis Test against PCA scores for suitable metadata fields.

{% if 'sigKruHeatmap' in item %}
Heatmap of Kruskal-Wallis Test against PCA scores for suitable metadata fields.
{% else %}

{{ item['Nkw'] }} fields where Kruskal-Wallis test between groups in PCA scores calculated.

{% endif %}

Figure 9. PCA scores plots coloured by metadata (significance by correlation).

{{ item['Ncorr_passing'] }} fields where correlation coefficient to PCA scores exceeded threshold of {{ item['r_threshold'] }}

{% if 'CORscores' in item %} {% for key, value in item['CORscores'].items() %}
PCA scores plots coloured by metadata (correlation).
{% endfor %} {% endif %}

Figure 10. PCA scores plots coloured by metadata (significance by Kruskal-Wallis).

{{ item['Nkw_passing'] }} fields where Kruskal-Wallis p-value against PCA scores exceeded threshold of {{ item['kw_threshold'] }}

{% if 'KWscores' in item %} {% for key, value in item['KWscores'].items() %}
PCA scores plots coloured by metadata (significance by Kruskal-Wallis).
{% endfor %} {% endif %}

Figure 11. PCA scores plots coloured by metadata (insufficent sample numbers to estimate significance).

{{ item['Nkw_insufficientClassNos'] }} fields where insufficent sample numbers to estimate significance.

{% if 'OTHERscores' in item %} {% for key, value in item['OTHERscores'].items() %}
PCA scores plots coloured by metadata (insufficent sample numbers to estimate significance).
{% endfor %} {% endif %} {% endblock %}