Abstract Recent technological innovations have significantly increased the efficiency of cryogenic electron microscopy (cryo-EM), causing a proliferation in the number of structures resolved with ever-increasing resolution. As many of the "low hanging fruit" structures have been determined, studies have shifted toward increasingly complex specimens, pushing the capabilities of current processing software and requiring novel methodologies. One class of challenging molecule for cryo-EM is tubular molecules that can have ambiguous symmetries. In particular, membrane remodeling proteins that form tubules are often only locally ordered making them intractable for structure determination. Another challenge for cryo-EM structure determination is the problem of the air/water interface. Often when blotting a sample for cryo-EM preparation, the sample can interact with the air/water interface causing protein denaturation, aggregation, or preferred orientations. We have developed a new technique that we are calling reconstruction of average subtracted tubular regions (RASTR) that has the potential to solve both of these problems. RASTR breaks down tubular cryo-EM samples into individual surfaces which enables structure determination without knowing or applying helical symmetry and classification of tubular samples that are only locally ordered with no long-range order. Here we propose two aims to make RASTR more robust, generalizable, and higher-resolution. These aims will be driven by challenging tubular samples that have proven to be intractable to typical cryo-EM structure determination methods. Finally, a third aim is to extend the RASTR approach to single particles bound to tubular substrates in order to protect them from the air/water interface and prevent preferred orientations. Together the proposed aims give RASTR the potential to become a general platform to enable structure determination of challenging samples.