How to split the Repetitive DNA file (rmsk) for human according to class and family using R

Question

[M K]

Hi,

I downloaded the Repetitive DNA file (rmsk) for human from UCSC website and I want to split this file according to class and family to get some basic statistics using R.

[Devon Ryan]

What have you tried and are you using the .out files or something from the table browser?

[M K]

Yes. I download this file from the following link

[Pierre Lindenbaum]

what kind of statistics, do you need the DNA sequences ?

[M K]

I want to get some basic statistics like frequencies for each family to and class to compare this file with mouse to see if there is any relation between them according to families and classes.

Answer 1

[Pierre Lindenbaum]

Using mysql ucsc (used mouse here, use hg19.simpleRepeat or hg19.nestedRepeat for human )

mysql --user=genome --host=genome-mysql.cse.ucsc.edu -A -D mm10 -e 'select repClass,repFamily,count(*) from rmsk group by 1,2'
+----------------+----------------+----------+
| repClass       | repFamily      | count(*) |
+----------------+----------------+----------+
| DNA            | DNA            |     1001 |
| DNA            | hAT            |     1949 |
| DNA            | hAT-Blackjack  |     4564 |
| DNA            | hAT-Charlie    |   105698 |
| DNA            | hAT-Tip100     |     9331 |
| DNA            | hAT-Tip100?    |      105 |
| DNA            | hAT?           |      585 |
| DNA            | MuDR           |      153 |
| DNA            | MULE-MuDR      |      583 |
| DNA            | PiggyBac       |      209 |
| DNA            | PiggyBac?      |      141 |
| DNA            | TcMar          |       52 |
| DNA            | TcMar-Mariner  |     1079 |
| DNA            | TcMar-Pogo     |       21 |
| DNA            | TcMar-Tc2      |     1786 |
| DNA            | TcMar-Tigger   |    35118 |
| DNA            | TcMar?         |      702 |
| DNA?           | DNA?           |     1027 |
| LINE           | CR1            |    14155 |
| LINE           | Dong-R4        |      138 |
| LINE           | L1             |   905176 |
| LINE           | L1?            |       52 |
| LINE           | L2             |    67909 |
| LINE           | RTE-BovB       |      260 |
| LINE           | RTE-X          |     1703 |
| LINE?          | Penelope?      |       42 |
| Low_complexity | Low_complexity |   386539 |
| LTR            | ERV1           |    71980 |
| LTR            | ERV1?          |      115 |
| LTR            | ERVK           |   319317 |
| LTR            | ERVK?          |     4185 |
| LTR            | ERVL           |   118061 |
| LTR            | ERVL-MaLR      |   454918 |
| LTR            | ERVL?          |      520 |
| LTR            | Gypsy          |     1859 |
| LTR            | Gypsy?         |      819 |
| LTR            | LTR            |      819 |
| LTR?           | LTR?           |      941 |
| Other          | Other          |    19450 |
| RC             | Helitron       |      345 |
| RC?            | Helitron?      |       74 |
| RNA            | RNA            |      691 |
| rRNA           | rRNA           |     1564 |
| Satellite      | centr          |        4 |
| Satellite      | Satellite      |    36865 |
| scRNA          | scRNA          |     8332 |
| Simple_repeat  | Simple_repeat  |  1015643 |
| SINE           | Alu            |   574557 |
| SINE           | B2             |   372923 |
| SINE           | B4             |   397726 |
| SINE           | Deu            |     1702 |
| SINE           | ID             |    64047 |
| SINE           | MIR            |   120436 |
| SINE           | tRNA           |     1618 |
| SINE?          | SINE?          |      274 |
| snRNA          | snRNA          |     3007 |
| srpRNA         | srpRNA         |      437 |
| tRNA           | tRNA           |     4769 |
| Unknown        | Unknown        |     6791 |
| Unknown        | Y-chromosome   |     2869 |
+----------------+----------------+----------+

[M K]

Thank, Pierre. I did that, but how can I split the rmsk file in R I need the information for each class and each family separately.

Answer 2

[Devon Ryan]

If you really want to use R (using mysql is actually faster), then you just want the split() command.

[M K]

I don't have experience for using mysql, also I want all information for each class and family like chr name, strand,...etc.