Modern LIMS have implemented extensive configurability as each laboratory's needs for tracking additional data points can vary widely. The LIMS may then import instrument results files to extract data for quality control assessment of the operation on the sample.

Modern targeted assays such as qPCR and deep sequencing can produce tens of thousands of data points per sample. Furthermore, in the case of drug and diagnostic development as many as 12 or more assays may be run for each sample. In order to track this data, a LIMS solution needs to be adaptable to many different assay formats at both the data layer and import creation layer, while maintaining a high level of overall performance. (source :http://en.wikipedia.org/wiki/Laboratory_information_management_system#LIMS_configurability)

Designing database based on the samples and the results depends upon following things/steps:

• Understanding file (samples or result) properties (name, author, date, location, no of sample)
• Designing Database Schema (tables) based on the properties of files (sample or result) and relationship
• Choosing Database Management tool (Mysql, postgresql, oracle, MongoDB etc)
• Programming language
• Designing parser to map needed file information into database.
• After mapping file content to database, Build GUI to query to the database.

These are the basic principles if you want to design database from files (sample or result)

Hope it help

Some people at the Bioinformatics Support Facility in my institute have developed a software to manage all the samples and experiment protocols, based on MediaWiki. It is called Lab Service Wiki. Maybe you can ask them if they are going to make their software freely available.

In any case, did you look at the other discussions in Biostar? There are many similar topics.