Free energy prediction of ligand binding to macromolecules using explicit solvent molecular dynamics (MD) simulations is computationally very expensive. Recently, we reported a linear correlation between the binding free energy obtained via umbrella sampling (US) versus the rupture force from steered molecular dynamics (SMD) simulations for epigallocatechin-3-gallate (EGCG) binding to α-helical-rich keratin. This linear correlation suggests a potential route for fast free energy predictions using SMD alone. In this work, the generality of the linear correlation is further tested for several ligands interacting with the α-helical motif of keratin. These molecules have significantly varying properties, i.e., octanol/water partition coefficient (log P), and/or overall charges (oleic acid, catechin, Fe(2+), citric acid, hydrogen citrate, dihydrogen citrate, and citrate). Using the constant loading rate of our previous study of the keratin-EGCG system, we observe that the linear correlation for keratin-EGCG can be extended to other uncharged molecules where interactions are governed by hydrogen bonds and/or a combination of hydrogen bonds and hydrophobic forces. For molecules where interactions with the keratin helix are governed primarily by electrostatics between charged molecules, a second, alternative linear correlation model is derived. While further investigations are needed to expand the molecular space and build a fully predictive model, the current approach represents a promising methodology for fast free energy predictions based on short SMD simulations (requiring picoseconds to nanoseconds of sampling) for defined biomolecular systems.