OpenEye FastROCS+  ·  3D Shape Similarity Report

Query: Gleevec (Imatinib) — Active Bound 3D Conformation  ·  Experiment TW-1 / 9052299
Query CompoundGleevec (Imatinib) · SPO-35646
Query ConformationActive Bound (Mol3D · CompoundPart 9004727)
MethodOpenEye FastROCS+ (Shape + Color Tanimoto)
Hits Returned54 Unique Compounds
Report DateMarch 18, 2026
Prepared ByT. Walker · twalker@sapiosciences.com
54
Total Hits Returned
Unique compounds after conformer collapse
0.755
Best Shape Tanimoto
Z494693834 — top 3D shape match
0.025
Lowest 2D Similarity
All hits are highly novel vs. Gleevec scaffold
10
Priority Follow-Up Compounds
High shape · Low 2D · IP-differentiated
Executive Summary

A 3D shape similarity search was performed using OpenEye FastROCS+ with the active bound conformation of Gleevec (Imatinib) as the query molecule. Gleevec's DFG-out inactive conformation was retrieved directly from the Mol3D field of CompoundPart record 9004727, ensuring the search probed for compounds that mimic the pharmacophoric 3D shape of Gleevec in its bioactive pose — the same geometry responsible for ABL kinase inhibition.

Key observation: All 54 returned hits display very low 2D Tanimoto similarity to Gleevec (range: 0.021 – 0.136), confirming they are structurally dissimilar at the 2D level despite overlapping well in 3D shape. This is precisely the profile sought for next-generation candidates that avoid Gleevec's intellectual property footprint while potentially retaining its binding geometry and pharmacological activity.

The top 10 compounds below were selected by maximising a Novelty Score defined as ShapeTanimoto − 2dTanimoto: a high value signals strong 3D shape complementarity to Gleevec's active conformation combined with maximum structural divergence from the competitor drug. These represent the most promising starting points for a lead optimisation campaign targeting ABL kinase with differentiated chemistry.

Shape vs. 2D Similarity — All 54 Hits
Ideal candidates cluster upper-left: high Shape Tanimoto (3D overlap), low 2D Tanimoto (structural novelty vs. Gleevec). Top-10 priority compounds are highlighted in orange. X-axis capped at 0.25 to show data spread.
Top-10 Priority Hits
Other Hits
Top 10 Priority Compounds — High Shape / Low 2D Similarity to Gleevec
Ranked by Novelty Score = ShapeTanimoto − 2dTanimoto. Higher score = better 3D mimic with greater structural novelty and IP differentiation.
# Compound ID Structure Shape Tanimoto 2D Tanimoto Novelty Score Combo Score Color Tanimoto Assessment
1 Z494693834 Z494693834
0.7551
0.0874
0.66780.96010.2050 Best Overall
2 Z909145572 Z909145572
0.7119
0.0577
0.65420.99050.3241 Lowest 2D
3 Z1410908738 Z1410908738
0.7373
0.0971
0.64021.03970.3024 High Shape
4 Z804255824 Z804255824
0.7127
0.0733
0.63940.95850.2458 Novel Scaffold
5 Z4743806970 Z4743806970
0.6626
0.0297
0.63290.87410.2210 Most Novel 2D
6 Z1974291854 Z1974291854
0.6892
0.0754
0.61380.94140.2523 High Shape
7 Z804171530 Z804171530
0.6815
0.0702
0.61130.99450.3131 Strong Shape
8 Z413351496 Z413351496
0.7241
0.1140
0.61011.19330.4692 Best Combo Score
9 Z90695798 Z90695798
0.6627
0.0600
0.60270.95000.2874 Strong Shape
10 Z4493815001 Z4493815001
0.6261
0.0251
0.60100.87220.2683 Lowest 2D in Top 10
Full FastROCS+ Results — All 54 Unique Compounds
Sorted by Shape Tanimoto (descending). Top-10 priority rows highlighted.
RankCompound IDShape Tanimoto 2D TanimotoNovelty ScoreTanimotoComboColor Tanimoto
Methodology & Scoring

Query structure: Gleevec (Imatinib, CAS 152459-95-5, SPO-35646) active 3D conformation retrieved from the Mol3D field of CompoundPart record 9004727 (MOL V2000, 37 atoms). The existing 3D coordinates were used directly — no new conformer generation was performed — ensuring the query reflects Gleevec's true pharmacophoric geometry in its DFG-out bound pose.


FastROCS+ scoring metrics:

  • Shape Tanimoto — volumetric 3D shape overlap with the query (0–1; higher = better shape match).
  • Color Tanimoto — electrostatic / pharmacophoric feature overlay (0–1).
  • TanimotoCombo — Shape + Color combined (0–2; practical threshold ≥0.8 = promising hit).
  • 2D Tanimoto — Morgan fingerprint (ECFP4) similarity to Gleevec. Low values indicate structural novelty and IP differentiation.

Novelty Score = ShapeTanimoto − 2dTanimoto. This metric rewards high 3D shape similarity while penalising structural closeness to Gleevec, prioritising differentiated chemical space for follow-up.