{% extends "base.html" %} {% block title %}Virtual Screening — MoleSight AI{% endblock %} {% block content %}
Target Selection
Filter by target: All Targets {% for t in targets %} {{ t }} {% endfor %}
Score vs QED Scatter
Compounds per Biological Target
Ranked Compound Library
{% for c in compounds %} {% endfor %}
Rank Name Target Score ↓ QED MW LogP IC₅₀ (nM) Drug-Like Activity
{% if loop.index0 == 0 %} #1 {% elif loop.index0 == 1 %} #2 {% elif loop.index0 == 2 %} #3 {% else %} #{{ loop.index }} {% endif %} {{ c.name }} {{ c.target }}
{{ c.score }}
{{ c.qed }} {{ c.mw }} {{ c.logp }} {% if c.ic50_nm < 10 %} {{ c.ic50_nm }} {% elif c.ic50_nm < 100 %} {{ c.ic50_nm }} {% else %} {{ c.ic50_nm }} {% endif %} {% if c.drug_like %} Yes {% else %} No {% endif %} {{ c.activity }}
Scoring Methodology
The composite screening score combines: pIC₅₀ (45%), QED drug-likeness (25%), LogP optimality (15%), TPSA (10%), and MW (5%), with penalties for Lipinski violations. This mimics a multi-parameter optimisation (MPO) approach used in real lead optimisation campaigns.
{% endblock %} {% block scripts %} {% endblock %}