Parsing API responses

When searches to the sample are parsed we initialize a new Sample object with the response as kwargs. This is easily breakable and can go unnoticed. We need to be able to improve testing here and make the parsing/initialization more robust.

Inspect the search response parsing and the Sample object initialization described in the issue. Identify the tests covering parsed search results, then add coverage for response data that does not match Sample initialization and make the parsing behavior explicit. Done means parsing is covered by tests and unexpected response fields cannot silently break initialization.

Written by the indexing model from the issue text.