Filtering rows for first two instances of a value

Shell Programming and Scripting
1

Kindly help me with this problem:

My data looks like this:

SNPfile.txt
CHR_A    BP_A    SNP_A    CHR_B    BP_B    SNP_B    R2    p-SNP_A     p-SNP_B
 4    172575323    rs17056855    4    172601079    rs11945883     0.119414    0.049972656    0.031050345
 4    172575323    rs17056855    4    172603701    rs7662060     0.11978    0.049972656    0.034664046
 4    172575323    rs17056855    4    172604186    rs17056919     0.382896    0.049972656    0.037136752
 4    172575323    rs17056855    4    172604350    rs4692884     0.11925    0.049972656    0.034632158
 4    172575323    rs17056855    4    172605017    rs17056925     0.373343    0.049972656    0.037902134
 5    148767855    rs12652986    5    148773386    rs353285     0.0727606    0.049972812    0.01748838
 5    148767855    rs12652986    5    148781294    rs353299     0.0692342    0.049972812    0.018621813
 5    148767855    rs12652986    5    148781964    rs353297     0.0324627    0.049972812    0.033439399
 5    148767855    rs12652986    5    148831342    rs414582     0.00996126    0.049972812    0.035560327
 5    148767855    rs12652986    5    148854047    rs10057083     0.00890368    0.049972812    0.03527856
 5    148767855    rs12652986    5    148866716    rs3756505     0.0106361    0.049972812    0.042519464
 21    19804594    rs7280435    21    19816817    rs456482     0.0613113    0.049972845    0.040731593
 21    19804594    rs7280435    21    19833893    rs128365    0.347322     0.049972845    0.01136657
 21    19804594    rs7280435    21    19865079    rs2226396     0.0193921    0.049972845    0.012638692
 21    19804594    rs7280435    21    19865343    rs2825653     0.0189789    0.049972845    0.018289668
 6    9690024    rs17797675    6    9703199    rs7749303    0.369563     0.049973276    0.018423278
 1    220966050    rs17532708    1    220970921    rs4240934     0.270007    0.049975232    0.010602066
 1    220966050    rs17532708    1    220972176    rs2378605     0.270007    0.049975232    0.010644871
 18    32233667    rs8092959    18    32264921    rs8087319     0.00315182    0.049975447    0.043287658
 12    2440796    rs4765937    12    2456906    rs2239087    0.416573     0.049978648    0.005179932
  1. I want to sort this first on col8 and then col9 in ascending order
  2. I also want to filter this list by taking the first two occurrence of row-value in col3 and write to another file with tab delimited format.

Output

CHR_A	BP_A	SNP_A	CHR_B	BP_B	SNP_B	R2	p-SNP_A	p-SNP_B
4	172575323	rs17056855	4	172601079	rs11945883	0.119414	0.049972656	0.031050345
4	172575323	rs17056855	4	172604350	rs4692884	0.11925	0.049972656	0.034632158
5	148767855	rs12652986	5	148773386	rs353285	0.0727606	0.049972812	0.01748838
5	148767855	rs12652986	5	148781294	rs353299	0.0692342	0.049972812	0.018621813
21	19804594	rs7280435	21	19833893	rs128365	0.347322	0.049972845	0.01136657
21	19804594	rs7280435	21	19865079	rs2226396	0.0193921	0.049972845	0.012638692
6	9690024	rs17797675	6	9703199	rs7749303	0.369563	0.049973276	0.018423278
1	220966050	rs17532708	1	220970921	rs4240934	0.270007	0.049975232	0.010602066
1	220966050	rs17532708	1	220972176	rs2378605	0.270007	0.049975232	0.010644871
18	32233667	rs8092959	18	32264921	rs8087319	0.00315182	0.049975447	0.043287658
12	2440796	rs4765937	12	2456906	rs2239087	0.416573	0.049978648	0.005179932

What I would have done if I wanted a unique list is to take only the col3 and then sort for uniq. However, I need two values per instance and I don't know how to do that.
In some cases, the values are present only once, adding to the problem.
I am looking for a solution in awk, since I am learning as it goes by posting some of the issues here.
I also want to thank many of you, who have helped me so far.

Sincere thanks
~GH

2 
sort -k8.1,8.10 -k9.1,9.10 infile > outfile

The number 2 requirement is not clear.
Does col 3 have to have repeating values, ie., 1 .. 1 in order to qualify as the first two?

This:

lends to the idea that somehow the contents of field #3 have to occur twice.

3

Posted the Output required.
If the Col3 occurs only once in some cases and then, it should also be written to the output. However, when Col3 value occurs more than once
like..

 CHR_A	BP_A	SNP_A	CHR_B	BP_B	SNP_B	R2	p-SNP_A	p-SNP_B
4	172575323	rs17056855	4	172601079	rs11945883	0.119414	0.049972656	0.031050345
4	172575323	rs17056855	4	172603701	rs7662060	0.11978	0.049972656	0.034664046
4	172575323	rs17056855	4	172604186	rs17056919	0.382896	0.049972656	0.037136752
4	172575323	rs17056855	4	172604350	rs4692884	0.11925	0.049972656	0.034632158
4	172575323	rs17056855	4	172605017	rs17056925	0.373343	0.049972656	0.037902134

It should write only the first two instances sorted on the columns p-SNP_A and p-SNP_B in that order.

CHR_A	BP_A	SNP_A	CHR_B	BP_B	SNP_B	R2	p-SNP_A	p-SNP_B
4	172575323	rs17056855	4	172601079	rs11945883	0.119414	0.049972656	0.031050345
4	172575323	rs17056855	4	172604350	rs4692884	0.11925	0.049972656	0.034632158

Also, can you explain how to sort a column with a mixture of values in decimals and scientific format (1E-06)?
Thanks
~GH

---------- Post updated 04-30-10 at 01:41 AM ---------- Previous update was 04-29-10 at 03:55 PM ----------

Please help!
vgersh my savior!!
Thanks
~GH