📊 Indirect Fourier Transform & the P(r) Function (SAXS Analysis)

This lecture section focuses on one of the most powerful tools in small-angle scattering (SAXS): the indirect Fourier transform (IFT) and the pair distance distribution function, P(r)

The key idea: We measure data in reciprocal space (I(q), intensity vs scattering vector q), but what we really want is real-space structural information about our macromolecule.

The Fourier transform is the mathematical bridge between those two spaces.

🔄 1. From Reciprocal Space to Real Space

In SAXS:

• I(q) → measured experimentally
• P(r) → real-space distance distribution inside the molecule

These two are mathematically connected through Fourier transform equations

You can:

• Transform I(q) → P(r)
• Or transform P(r) → I(q)

When you transform P(r) back into I(q), you get a smooth fitted curve to your experimental data.

If that smooth line matches your experimental points well → your transform worked properly.

🧮 2. What Is the P(r) Function?

The P(r) function (pair distance distribution function) is:

A histogram of all pairwise electron–electron distances in the macromolecule

Imagine your protein:

• Take every pair of electrons
• Measure the distance between them
• Count how many pairs occur at each distance
• Plot:
  • x-axis = distance (r)
  • y-axis = number of electron pairs at that distance

That gives you P(r).

It is literally a structural fingerprint of the molecule.

📏 3. Maximum Dimension (Dmax)

One immediate feature:

• P(r) always goes to zero at some maximum distance.
• That distance is the maximum dimension of the molecule (Dmax)

Why?

Because beyond that distance, there are no more electron pairs inside the object.

So:

D_ = ext{largest internal distance in the molecule}

This is extremely important for structural modeling.

🔵 4. How Shape Affects the P(r) Curve

The shape of the P(r) curve depends strongly on molecular geometry

Let’s go through the major cases.

🟠 (A) Solid Sphere (Globular Protein)

Shape of P(r):

• Symmetric
• Almost Gaussian
• Smooth rise and fall

This is what most globular proteins look like.

Interpretation:

• Many electron pairs at intermediate distances
• Fewer at very short and very long distances

If your protein is compact and folded → expect this shape.

🟢 (B) Long Rod

Features:

• Sharp peak at low distances
• Long tail extending toward Dmax

Why?

• Many electron pairs exist across the short width (short r)
• Fewer but important distances span the long axis (large r)

This produces:

• Early strong peak
• Extended tail

Common for:

• Fibrous proteins
• Elongated complexes

🟣 (C) Disc

Looks somewhat similar to sphere but:

• Broader distribution
• Peak occurs earlier

Because it’s flatter, distances are distributed differently.

🟡 (D) Hollow Sphere

Opposite behavior of rod:

• Large number of long distances
• Peak near the maximum dimension

Why?

Most electrons are arranged in a shell → many distances span the entire diameter.

🔵 (E) Dumbbell (Two Domains)

This is extremely important biologically.

Features:

• First peak = distances within each domain
• Second peak = distances between domains

This is typical for:

• Multi-domain proteins
• Proteins with flexible linkers

The second peak corresponds to inter-domain spacing.

If you see two peaks → think domain organization.

🧬 5. Real Protein Examples

From actual SAXS data :

Globular proteins

• Similar to sphere
• Slight tail

Multi-domain proteins

• Shoulders or secondary peaks
• Inter-domain distances visible

Unfolded proteins

• Compressed at short distances
• Very long extended tail

Unfolded systems show much more extended distributions.

This becomes important when studying:

• Protein flexibility
• Folding
• Disorder

📐 6. What Can You Extract from P(r)?

The P(r) function gives:

✅ 1. Maximum dimension (Dmax)

Clear cutoff where curve goes to zero.

✅ 2. Radius of gyration (Rg)

You can calculate Rg directly from P(r).

This can be:

• More accurate than Guinier analysis
• Especially useful for:
  • Large particles
  • Noisy data
  • Small Guinier range

Because P(r) uses the entire curve, not just low-q points.

✅ 3. I(0) (Forward scattering intensity)

Can also be obtained from P(r).

Good for cross-checking:

If:

• Guinier Rg ≈ P(r) Rg
• Guinier I(0) ≈ P(r) I(0)

Then your data processing is likely reliable.

⚠️ 7. Sensitivity to Problems

The P(r) function is sensitive to:

• Aggregation
• Interparticle interference

If your P(r):

• Doesn’t smoothly go to zero
• Shows strange oscillations
• Has unexpected long tails

→ something may be wrong with the sample.

This makes P(r) a powerful diagnostic tool.

🧠 8. Why Is Indirect Fourier Transform Necessary?

It is:

• Model-independent
• Real-space based
• Required before advanced modeling

Especially important because:

• Dmax is needed for ab initio shape reconstruction
• It constrains the search space
• It improves reliability of structural modeling

Without a proper P(r), you cannot confidently move forward to 3D reconstructions.

📌 9. Big Picture Summary

Indirect Fourier Transform allows you to:

🔄 Convert reciprocal space data (I(q)) ➡ Into real-space structural information (P(r))

P(r) tells you:

• Molecular shape
• Maximum dimension (Dmax)
• Radius of gyration (Rg)
• I(0)
• Presence of multiple domains
• Folding state
• Flexibility
• Aggregation artifacts

It is:

• Model-independent
• Highly informative
• Required for advanced analysis
• More robust than Guinier in many cases

🧩 Conceptual Takeaway

Think of I(q) as:

A blurry fingerprint in reciprocal space.

And P(r) as:

The real-space histogram of all internal distances — the molecule describing itself from the inside.

The transform is simply the mathematical bridge between those two worlds.