Spider dragline silk is a self-assembling tunable protein composite fiber that rivals many engineering fibers in tensile strength, extensibility, and toughness, making it one of the most versatile biocompatible materials and most inviting for synthetic mimicry. While experimental studies have shown that the peptide sequence and molecular structure of silk have a direct influence on the stiffness, toughness, and failure strength of silk, few molecular-level analyses of the nanostructure of silk assemblies, in particular, under variations of genetic sequences have been reported. In this study, atomistic-level structures of wildtype as well as modified MaSp1 protein from the Nephila clavipes spider dragline silk sequences, obtained using an in silico approach based on replica exchange molecular dynamics and explicit water molecular dynamics, are subjected to simulated nanomechanical testing using different force-control loading conditions including stretch, pull-out, and peel. The authors have explored the effects of the poly-alanine length of the N. clavipes MaSp1 peptide sequence and identify differences in nanomechanical loading conditions on the behavior of a unit cell of 15 strands with 840–990 total residues used to represent a cross-linking β-sheet crystal node in the network within a fibril of the dragline silk thread. The specific loading condition used, representing concepts derived from the protein network connectivity at larger scales, have a significant effect on the mechanical behavior. Our analysis incorporates stretching, pull-out, and peel testing to connect biochemical features to mechanical behavior. The method used in this study could find broad applications in de novo design of silk-like tunable materials for an array of applications. © 2011 Wiley Periodicals, Inc. Biopolymers 97: 408–417, 2012.