Determining structure from a single amino acid change, let alone an insertion, is not a trivial problem. Structural modeling programs such as Rosetta (https://www.rosettacommons.org/) or IMP (https://integrativemodeling.org/) can give you a good start if the protein already has a known structure in the PDB database. However the new structure is still a hypothesis, and it would require the structure to be experimentally resolved using crystallography, CryoEM, NMR, etc. All of these methods require expertise in either simulation or experimental design, and would probably be best done by a collaborator who is knowledgable in these spaces.

You could try using SIFT, PolyPhen 2 online tools to predict changes in structure/function due to mutation. Performing an MD simulation of the wild type/mutant protein can provide significant insights on the effect of mutation on the protein structure.

It depends very much on where in the protein such a change occurs. If it severely truncates a domain or changes a majority of the domain sequence, you can be pretty sure that this domain no longer will fold properly. Domains N_terminal to the affected domain may still properly fold and even be functional. This could be a base for dominant negative effects, if such a protein e.g is still able to form heterodimers with the full length wild-type protein due to the intact N-terminal domains, but is no longer able to trigger effector functions through the domain affected by the mutation. In such a case, the mutated protein would compete with the w.t. protein. Think, for example, of a tyrosine kinase receptor with a truncation in the kinase or autophosphorylation domain. It the truncation lies beyond the transmembrane domain, the extracellular domains can show ligand-dependent dimerisation, but if only one of the two monomers contains a functional kinase, this dimerisation cannot lead to the kinase dimerisation needed for activation.