Using Deep Learning in NGS

Using Deep Learning in NGS.

I’m trying to study for my Computer Science course and I need some help to understand this question.

-You have recently taken a job as a bioinformatician in an NHS clinical genomics unit.

-The focus of the group is the use of NGS technology for patient care Within the unit the bioinformatics resources are relatively small (2 or 3 people) in a team of 7 or 8 – and you are the only person in the team with a computer science background.

-For one of your first tasks, the head of unit has asked you to produce a short report on an informatics problem being faced. This report should have three sections (about 6 pages in total), and an appendix.

-Section1:An analysis of the problem, defining its scope

-Section 2: A discussion of the strategies that could be followed to address the problem

-Section 3: A recommendation of the strategy to be followed.

-It is important that the report be accessible to someone who is scientifically literate, but not an expert computing or information systems.

Deep learning. The quantity of data available on human variants has massively expanded in the past year. It is expected that the amount data will increase even further (5 million genome project). The lab head has been to a number of talks which are highlighting the role that “deep learning” methods could play in interpreting genomics data in the context of disease. Should the lab begin to explore deep learning in its own work?

Questions:

What is deep learning?

Is it mature enough to deliver real benefit in a clinical setting?

Section

0-4

5-6

6-7

7-10

Scenario background research

Includes minimal relevant information – contains many errors or much of what is included is irrelevant to the topic

Covers several issues, but these are listed rather than linked.

Several issues are discussed and related to the problem.

Covers all the major areas, places them in the wider context, and shows a degree of originality or professionalism. Typically this is work that would be considered of publishable standard.

Strategy

Strategy applied contains numerous errors and largely inappropriate

Strategy is basically correct. Some of the interpretation might be weak – some issues in developing a convincing and supportable strategy.

A number of strategies are properly discussed that cover the expected areas.

An excellent set of strategies that uses appropriate tools, develop properly supported ideas, and shows a good understanding of the underlying issue

Conclusions

Little or no support for the conclusion reached

An attempt is made to support the final conclusion, but it is flawed in places.

A good discussion of the strategies to support the final conclusion.

An excellent discussion of the strategies to support the final conclusion. The recommended strategy is of professional quality.

Report quality

Numerous errors with little referencing. Poorly structured or disorganised

Report has a good basic structure, minimal careless errors and appropriate presentation

A well written report that is well structured and with good referencing.

A report that is written and presented to a very high standard. Good use is made of figures and tables. Referencing complete and appropriate.

Using Deep Learning in NGS

"Looking for a Similar Assignment? Order now and Get a Discount!