"These two molecules look almost the same..."
Sena showed two structural formulas side by side.
"But the activity?" Akira asked.
"One has IC50 of 10 nM, the other 1 μM. A 100-fold difference."
Mikhail quietly approached. "Similar but not the same, they say."
"What's different?"
Akira looked at the structures carefully. "Pyridine and benzene, here's the difference."
"But the shapes are almost the same, right?"
"Shape is similar. But properties differ," Mikhail explained. "Pyridine has nitrogen. Basicity, hydrogen bond acceptance, dipole moment... all different."
Akira continued. "This nitrogen makes an important interaction with somewhere in the active site."
"Hydrogen bonding...?"
"High possibility. If there's a donor on the protein side, pyridine's nitrogen works as an acceptor."
Mikhail rotated the structure. "But benzene can't make that interaction."
"So activity drops..."
Akira drew on the whiteboard. "This is called a pharmacophore. Structural features essential for efficacy."
"Pharmacophore?"
"Hydrogen bond donors, acceptors, hydrophobic centers, aromatic rings... the 3D arrangement of these determines activity."
Mikhail supplemented. "With the same pharmacophore, completely different scaffolds can work the same way."
"Even different scaffolds?" Sena was surprised.
"It's a strategy called scaffold hopping," Akira explained. "Change the basic skeleton to avoid patents or improve properties."
Mikhail showed an example. "This and this have completely different scaffolds but equivalent activity."
"Indeed... but the pharmacophore is in the same position."
"Yes. From the protein's perspective, both look the same."
Akira introduced another concept. "Bioisosteres are also important."
"Bioisostere?"
"Biologically equivalent substituents. For example, carboxylic acid and tetrazole."
Mikhail drew the structures. "Shape, size, and pKa are similar. So activity is often maintained when substituted."
"But not perfectly the same..."
"That's what's interesting," Akira said. "Because properties differ subtly, metabolic stability and membrane permeability can improve."
Sena took notes. "So it's important to discern similar and different parts."
"Exactly," Mikhail acknowledged. "Distinguishing what's essential and what's additional is where a designer shows their skill."
Akira showed other data. "This is similarity search results. Calculating similarity with Tanimoto coefficient."
"Above 0.8 means structurally similar."
"But even at 0.9, activity can be completely different," Sena pointed out.
"Conversely, even around 0.5, they sometimes work the same way," Mikhail added.
"Can't tell from numbers alone..."
"Chemical intuition is needed," Akira said. "Learn from experience which parts are involved in interactions."
Mikhail said with a serious expression, "But there are oversights. Judging with preconceptions makes you ignore important structures."
"So what should we do?"
"Look at data carefully," Akira answered. "Create series with gradual structural changes and identify what works."
"SAR basics," Mikhail continued. "Change one thing at a time and measure each contribution."
Sena pondered. "It's painstaking work."
"But it's a reliable method," Akira smiled. "Steady accumulation is more important than flashy discoveries."
Mikhail opened another screen. "This is similarity predicted by machine learning. Learning from activity patterns, not just structure."
"Even if structures differ, if activity patterns are similar, they're classified in the same class."
"Interesting approach."
"But interpretation is still difficult," Mikhail said. "It tends to be a black box."
Akira summarized. "To discern similar but different requires looking comprehensively at structure, properties, and activity."
"Can't see the essence from just one aspect..."
"Yes. Having multifaceted perspectives. That's a good designer's requirement."
Outside the window, the sun was setting. Similar yet different. Different yet similar. The eye to discern that boundary was being cultivated again today.
"Next, let's try 3D pharmacophore modeling," Akira suggested.
"Thinking in three dimensions."
"Yes. Features invisible in planar structures become visible in 3D."
Sena's heart swelled with anticipation. The world of similar but different seemed even deeper.