Question

Ok let’s first look at a screen shot. This is a screen shot of a text file we call it VCF file. How many rows it might have? Maybe 100,000 rows of things like this:

I am totally new and novice to MongoDB so I thought of a schema like this:

So for example notice REF in that text file is a Key/Value in my schema. But like I said it might have 200,000 rows…
So:

1. Are Arrays still a good thing I can use? storing 200,000 members in that array?
2. How powerful I can query on it? so in the text file we have rows, for example that #CHROM20 in POS of 14370 has a REF of “G” and ALT of “A” … so with my Schema can we find and return it? Let’s say we say search for patients that have “G” in their REF field, so are MongoDB queries powerful enough to search and return such a result?
3. Is it a bad schema? Do you have better recommendations/advice?
4. Any sample query could you give for my qquesry in question will be so helpful to give me some ideas..

Answer 1

Sorry for the very slow reply, I had left for holiday when you replied. The following syntax achieves the desired outcome.

> db.refs.insert({ref:[A,T,ATC,G]})
> db.refs.insert({ref:['A','T','ATC','G']})

> db.refs.findOne()
{
    "_id" : ObjectId("4fda21bb8a807d87a65aba37"),
    "ref" : [
        "A",
        "T",
        "ATC",
        "G"
    ]
}
> db.refs.insert({ref:['TCG','TA']})
> db.refs.find()
{ "_id" : ObjectId("4fda21bb8a807d87a65aba37"), "ref" : [ "A", "T", "ATC", "G" ] }
{ "_id" : ObjectId("4fda22438a807d87a65aba38"), "ref" : [ "TCG", "TA" ] }


> db.refs.find({ref :{$all : ['G']}})
{ "_id" : ObjectId("4fda21bb8a807d87a65aba37"), "ref" : [ "A", "T", "ATC", "G" ] }

Is this what you had in mind?

A big concern in schema design is avoid the 16MB document limit. While you can have as many documents as can be addressed with 64 bit address space, I don’t know how your document is likely to grow. This restriction may necessitate that you break out some of the fields into other documents that you reference.