A quick recipe using Linux Shell and BEDtools. Personally I much prefer R and GenomicRanges for this workflow, but this R-based method is not the place to start for newbies.

I've left some of the technical detail out as it's a chance to teach yourself some basic bioinformatics. Step 15 is most important!

Recipe:

1) Go to http://genome.ucsc.edu/cgi-bin/hgTables

2) Select 'Mapping and Sequencing' group, 'Chromosome Band' track and 'cytoBand' table.

3) Select 'All fields from selected table' as the output format.

4) Save the output as 'cytoBand.txt'

5) Now select 'BED' as the output format.

6) Save the output as 'cytoBand.bed'

7) Get your hands on a Linux based (or maybe MacOSX) workstation (seriously, don't try and do bioinformatics with Windows! It's like trying to tie shoelaces with one hand, while hopping).

8) Filter the table using shell commands on the gieStain column using grep and file redirection. e.g. for gneg only rows, "cat cytoBand.txt | grep gneg$ > cytoBand_gneg.txt"

9) Convert the file to a legal BED file using the shell 'cut' command (http://www.thegeekstuff.com/2013/06/cut-command-examples/) or write a Perl or Python script.

10) Check your new file looks much like the original BED file you downloaded (except with less rows).

11) Download BEDtools, https://github.com/arq5x/bedtools2

12) Learn about how to intersect your BAM file with the BED file here: http://bedtools.readthedocs.org/en/latest/

13) Create subset BAM files using BEDtools.

14) Save scripts, logs and document your work for the sake of reproducible research!

15) Teach your new bioinformatics skills to other lab peeps!!

Yes, of course. You can use Giemsa banding technique.

Heterochromatin is full of repeats, therefore you will have troubles to map these reads uniquely to the genome. You can generate two baskets, one with uniquely mapping reads and one with multiple mappings, representing euchromatin and heterochromatin respectively.

If you have paired end reads, more will map in heterochromatin. However, pericentromeric, centromeric and telomeric reads will still be under-represented as uniquely mapping reads.

To filter heterochromatic-region derived reads, use the subsetByOverlap() function in GenomicRanges in R/Bioconductor, or BEDTools in the shell to filter against the UCSC cytoBand table. You may want to grep this table first to remove all rows with Giemsa staining that doesn't match your required filter.

If none of that bioinformatics workflow above makes sense, watch some GenomicRanges tutorials and Google up some BioStars forum posts, or make friends with a local bioinformatician.

Dear Herbert,

Thanks for the information.

It make sense let me try it out

Dear Jason,

Thanks for the clear description.

It looks bit complicated for me since I am new to this bioinformatics. But your information looks interesting.

I will try to do and get back to you after that.

Thanks

A quick recipe using Linux Shell and BEDtools. Personally I much prefer R and GenomicRanges for this workflow, but this R-based method is not the place to start for newbies.

I've left some of the technical detail out as it's a chance to teach yourself some basic bioinformatics. Step 15 is most important!

Recipe:

1) Go to http://genome.ucsc.edu/cgi-bin/hgTables

2) Select 'Mapping and Sequencing' group, 'Chromosome Band' track and 'cytoBand' table.

3) Select 'All fields from selected table' as the output format.

4) Save the output as 'cytoBand.txt'

5) Now select 'BED' as the output format.

6) Save the output as 'cytoBand.bed'

7) Get your hands on a Linux based (or maybe MacOSX) workstation (seriously, don't try and do bioinformatics with Windows! It's like trying to tie shoelaces with one hand, while hopping).

8) Filter the table using shell commands on the gieStain column using grep and file redirection. e.g. for gneg only rows, "cat cytoBand.txt | grep gneg$ > cytoBand_gneg.txt"

9) Convert the file to a legal BED file using the shell 'cut' command (http://www.thegeekstuff.com/2013/06/cut-command-examples/) or write a Perl or Python script.

10) Check your new file looks much like the original BED file you downloaded (except with less rows).

11) Download BEDtools, https://github.com/arq5x/bedtools2

12) Learn about how to intersect your BAM file with the BED file here: http://bedtools.readthedocs.org/en/latest/

13) Create subset BAM files using BEDtools.

14) Save scripts, logs and document your work for the sake of reproducible research!

15) Teach your new bioinformatics skills to other lab peeps!!

Dear Jason,

Thank you so much for your step by step explanation.

It really helps a lot.

Thanks,

Kumar..

Heterochromatin is naturally condensed and had the tendency to contain multiple repeats and frequently non-expressed or inactive sequences. On the other hand, euchromatin is less condensed, has fewer repeats and mostly active and expressed sequences. Giemsa staining can clearly be used to map these regions physically on chromosomes. If the spread is well prepared, giemsa staining can be removed especially since the euchromatin that stains deeply, has low condensation. In situ hybridization can be applied to map other sequences physically on the same spread if need be.  

some suggestions : heterochromatin is usually associated with the nuclear enveloppe, you could try to find your candidate sequences associated to these membranes ; histone modifications are distinctive between euchromatin and heterochromatin, again you could try to selectively isolate your candidate sequences with antibodies against these specific modified histones

Dear Herve,

Thanks again for the update..

Hi Kumar,

Just wanted to expand on everyone else's suggestions in case it might help.

This kit is very useful if you're looking to assess euchromatin or heterochromatin states for just a few genes: http://www.epigentek.com/catalog/epiquik-chromatin-accessibility-assay-kit-p-3564.html 

http://www.epigentek.com/catalog/epiquik-chromatin-accessibility-assay-kit-p-3564.html