Using mysql/ucsc:

$ mysql --user=genome --host=genome-mysql.cse.ucsc.edu -A -D hg19 -e 'select X.geneSymbol,K.name as transcript, K.chrom,K.txEnd-K.txStart as transcriptLen from kgXref as X,knownGene as K where X.geneSymbol="KCNQ2" and X.kgId=K.name '
+------------+------------+-------+---------------+
| geneSymbol | transcript | chrom | transcriptLen |
+------------+------------+-------+---------------+
| KCNQ2      | uc002yex.3 | chr20 |          1804 |
| KCNQ2      | uc002yey.1 | chr20 |         66452 |
| KCNQ2      | uc002yez.1 | chr20 |         66452 |
| KCNQ2      | uc002yfa.1 | chr20 |         66452 |
| KCNQ2      | uc002yfb.1 | chr20 |         66452 |
| KCNQ2      | uc002yfc.1 | chr20 |         38962 |
| KCNQ2      | uc011aax.1 | chr20 |         48707 |
+------------+------------+-------+---------------+

You might look up the RefSeq accession and version name:

$ wget ftp://ftp.ncbi.nlm.nih.gov/refseq/H_sapiens/RefSeqGene/gene_RefSeqGene
...
$ grep 'KCNQ2' gene_RefSeqGene
9606    3785    KCNQ2   NG_009004.1

Then look up the alignment:

$ wget ftp://ftp.ncbi.nlm.nih.gov/refseq/H_sapiens/RefSeqGene/GCF_000001405.26_refseqgene_alignments.gff3
...
$ grep 'NG_009004.1' GCF_000001405.26_refseqgene_alignments.gff3
NC_000020.11    RefSeq  match   63404189        63477640        73452   +       .       ID=aln1679;Target=NG_009004.1 1 73452 -;bit_score=136701;e_value=0;gap_count=0;num_ident=73452;num_mismatch=0;pct_coverage=100;pct_identity_gap=100;pct_identity_gapopen_only=100;pct_identity_ungap=100;score=73452

The accession NC_000020.11 tells you KCNQ2 is mapped to chromosome 20 at bases 63404189 to 63477640. Note that this alignment is to assembly hg38, which you may or may not be using:

https://genome.ucsc.edu/cgi-bin/hgTracks?db=hg38&position=chr20%3A63404189-63477640

I've not seen a convenient database of this (this isn't usually stored explicitly), but you can determine this given an annotation file easily enough. Here's an example of how to do that (and a bit more) in R: Normalization Of Rna Sequencing Counts (By Ercc / Gene Length)

A: Length of transcripts of mm9

This was a long, complicated, and tiring process. Getting the variant density is annoying as well if you don't know how to use R very efficiently. I'm just about done with it. If I'm allowed to later, I will post the methods up. However, if you don't care about distinguishing regions that are intron/exon/UTR/coding/etc., then just look at the gtf. Extract the whole genes. If you want a specific region, you'll have to extract that type and concatenate the results from the gtf. If you're dealing with targeted sequencing, then the process is more complicated.