marineLM / NetNorM

6 stars 4 forks source link

Interpreting output files #1

Open code837 opened 6 years ago

code837 commented 6 years ago

Hi -

The netnorm program was utilized by my team members recently, and they used LUAD data from TCGA. We are wondering what to do with the netnorm output (do we feed it to a visualizer, etc). Do you recommend any programs that we can feed this output to?

Here are some snippets from the output:

FILE: survpred_LUAD_mut_PC_neighbours_stair_rsfolds4_False_rsrandNA_50_10_375_NA_NA_NA CONTENTS:

(lp0 cnumpy.core.multiarray _reconstruct p1 (cnumpy ndarray p2 (I0 tp3 S'b' p4 tp5 Rp6 (I1 (I5 tp7 cnumpy dtype p8 (S'f8' p9 I0 I1 tp10 Rp11 (I3 S'<' p12 NNNI-1 I-1 I0 tp13 bI00 S'\x10\xefl\xedE\x80\xee?i-? CR\xef?l\x00]IU\xa0\xe9?\xe6\xf4\x9f\xc6\x15P\xef?\x02\xe0\x05\x8b\xcaF\xef?' p14 tp15 bag1 (g2 (I0 tp16 g4 tp17 Rp18 (I1 (I5 tp19 g11 I00 S'\xe6\x96\xfd\x86x\xa2\xe5?\xfep#\xa4 \xe2?|\xd8\xfa6\xa8\xe2?\xa9\xd6\xd9\x17Q\xad\xe3?&N\xfd\x9e[\xe2\xe4?' p20 tp21 ba(lp22 g1 (g2 (I0 tp23 g4 tp24 Rp25 (I1 (I75 tp26 g8 (S'O8' p27 I0 I1 tp28 Rp29 (I3 S'|' p30 NNNI-1 I-1 I63 tp31 bI00 (lp32 S'ADAM19' p33 aS'ADCY2' p34 aS'ADNP2' p35 aS'AIFM1' p36 aS'ALYREF' p37 aS'ANK2' p38 aS'ANXA1' p39 aS'ANXA2' p40 aS'BCL11B' p41 aS'CDH7' p42 aS'CDK1' p43 aS'COPA' p44 aS'CRB1' p45 aS'CSNK1A1' p46 aS'CSNK2A2' p47 aS'DDX17' p48 aS'EGFR' p49 aS'FAM5C' p50 aS'FLNB' p51 aS'FN1' p52 aS'GNAI2' p53 aS'HADHB'

(I1 (I16674 tp319 g11 I00 S'\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x0

(I86 tp435 g11 I00 S'\x98\xa6\xf6RI(\xbb\xbf\xc3*\x8bM\xae\xdf\xbe\xbf\xe4\xb0<sD.\xe7\xbf^\xd9\x06\xe8\xc2\x19\xcc?l\x9c\xcf\x12\x1f\xee\xe6\xbf \xf7L\xd7\xb3\xfd\xd3?X\x1d\x18\x9c\x18D\xe4\xbf\x00\x9b\xc4\x96\x9c\x1a\xb4\xbfXE\xa2\x1fS\x92\xbf+9\xb0\x08b\xb1\xcc\xbf\xf8\xad\xb1\x1fi\x14\xdb\xbf\\\xf9\xa9\xf3\x8b+\xc7\xbf\x00lL\x84\xcc)8?\xea\x8evx\xf4\xe3\xe0\xbf\x15H\xd7s\xe3\xc6\xc2\xbf4\x0c\x0c/\xd6(\xb9\xbf%\xc2\x8dP\x0fn\xc6\xbf\xe4\xa7\xf6\x8f\x0bE\xd8\xbf\x96\x142\x0b\xb2!\xd5\xbf\x8aY\x19\xc0\xa0\x1e\xe2\xbf\x01W\n\x9fs\x9c\xc9\xbf<L\xeb\xeb\x05e\xe6\xbf-\x9b\x8a\xa8-\x90\xc9\xbf\'\xf2\'\x8b\xf2z\xc5\xbf\x7f\x02\x87\xdd^\xf9\xd1\xbf\xb8,\x17\xbf\x90U\xec\xbfT\xed\xd0L\xc8\xdf\xce\xbf\x1c\x8a\xf3f\xc2\x9b\xd8\xbf];\x9f\x99\xaeN\xc9\xbf\x96\xf6\\\x01\xfa\xbc\xe3\xbf\x02\x9e7C\xc4\xd3\xd3\xbfs>\x98\x08!\xa9\xd2\xbf\x10\x07\xeb\xf12$\x90?I\xb2\n\x81^:\xdf\xbf\x1a\x1aY\xac\xb1,\xd3\xbf|\x08\xf3G|#\xa1?3\x9f\xd8!\x13Q\xd0?T\x10\xfbY\xf2\xb7\xbd\xbf83\xf4r\xe3h\x95\xbfb\x005.?b\xd3\xbf1\xea\xe9\xa4\xbc\xe1\xc4?\xf0\xa7\xf2,8r\xe6\xbft\xeb\xed\xb6\x16\xe0\xdd\xbf\xb3\xe4d\xcf\xdf\xe7\xd0?\xecZ\xae\xdd\xd7\xc4\xde\xbf\xa8ce\xde\x04\xca\xbf\x14\xbe\x88\x1d\xc4D\xeb?8\xa4\xcd\xfb\xb3&\xdf\xbf\x05\x8d\xce\xb1\xb8Q\xd9\xbfe)\xc3{/\xc4\xc0\xbf\xe2\xdds\x05\x93\x02\xe0\xbff\xdc\xc8\xfdz\x8f\xbb?\x96A|h\xb0\x8c\xd7\xbf\x1c\x1d\xd8Y:\x9f\xe4\xbf\xe2\xe9m\x9e\xaa\x0e\xe4\xbf\xe6\xd1\x9aL\x81\xee\xe7\xbf\xf4\xa0\xe0d\x88\xd1\xcd?2\xad\xcc\xf4\xec\xcf\xaf\xbfl\x7fgh\xf7\xba\xe2\xbfPNi\xe3\xd0\x1d\xd0?\xa4\x0c\x13\xa4\xf2%\xcd\xbf\x85&\xf8\xea\x04\xa3\xce\xbf\x9eqHJ\x80;\xf0\xbf3\x03\xc9\n<\x08\xd8\xbf-z\xa5\x8e\x7fK\xd1\xbf\x82\x119\xf2\x8f>\xb7?\xcc\x05\x91\x93#<\xe3\xbf\xa4\xb6kK\xf4\xf0\xd0?\x80\xab&9\x81\xa2\x98\xbf\xd8\xf3n\x84\U\xd5\xbf\xddOx\xd9-\x1c\xf1\xbfnH3\x1cZ\x08\xdc\xbf\x12e\x86\xf9#\xf0\xda\xbf\xce\xdb\xf6)\x0b<\xce\xbf\xfb\xfc\x17\xd7\x11_\xea\xbf\xaf>RW}\xc2\xd2\xbf \x96f\xacx\x96\xd6\xbfl\xdc\x01\xf6\xfd\xe4\xd0\xbf\xdf\\z\xeb\x8a{\xbfH{\x9c\x95\xfe\xb4\x9e\xbf\xc3@\x9f\x7f\x06\x1c\xb7\xbfOS2\x0fB\xc5?B\xdf\xca\x10\xe7\xd2\xa2\xbf\xb9\xef\xf6YAk\xd9\xbfz4[\xc1(\xf3\xe4\xbf\xbe\x8e~\n"f\xf1\xbf' p436 tp437 baa.

FILE: stratNMF_LUAD_PC_neighbours_stair_k295_alphaNA_randomizeFalse_rs_randNA_N5 CONTENTS:

ccopy_reg _reconstructor p0 (cpandas.core.series Series p1 cbuiltin object p2 Ntp3 Rp4 (dp5 S'_metadata' p6 (lp7 S'name' p8 asS'_data' p9 g0 (cpandas.core.internals SingleBlockManager p10 g2 Ntp11 Rp12 ((lp13 cpandas.core.indexes.base _new_Index p14 (cpandas.core.indexes.base Index p15 (dp16 S'data' p17 cnumpy.core.multiarray _reconstruct p18 (cnumpy ndarray p19 (I0 tp20 S'b'

(lp31 S'TCGA-05-4244-01' p32 aS'TCGA-05-4249-01' p33 aS'TCGA-05-4250-01' p34 aS'TCGA-05-4382-01' p35 aS'TCGA-05-4384-01' p36 aS'TCGA-05-4389-01' p37 aS'TCGA-05-4390-01' p38 aS'TCGA-05-4395-01' p39 aS'TCGA-05-4396-01' p40 aS'TCGA-05-4397-01'

NNNI-1 I-1 I0 tp474 bI00 S'\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x01\x00\x00\x00\x00\x00\x00\x00\x04\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x01\x00\x00\x00\x00\x00\x00\x00\x

ag456 ag457 ag458 ag459 ag460 ag461 atp484 bsg8 Nstp485 Rp486 a(dp487 S'0.14.1' p488 (dp489 S'axes' p490 g13 sS'blocks'

marineLM commented 6 years ago

Hi,

Running the function "survpred" gives you an output which is a python list. This list is described as follows:

-------
l: list
    If 'data_type' is set to 'clin' or 'mut', the list contains 12
    elements. Each of these elements is itself a list or an array of
    length 5 corresponding to the 5 splits in train/test of the
    cross-validation. These containers of length 5 record:
    #0: Concordance index (CI) obtained on the train set.
    #1: Concordance index (CI) obtained on the test set.
    #2: the variables selected in the model (genes or clinical features)
    #3: the parameters ((k and C) or (alpha and C) according to the
        method used) that was learned with the inner cross-validation.
    #4: the coefficients learned for each variable.
    #5: Concordance index (CI) obtained on the train set for patients with
        more than the mean number of mutations.
    #6: Concordance index (CI) obtained on the test set for patients with
        more than the mean number of mutations.
    #7: Concordance index (CI) obtained on the train set for patients with
        less than the mean number of mutations.
    #8: Concordance index (CI) obtained on the test set for patients with
        less than the mean number of mutations.
    #9: Mean concordance index (CI) obtained across in the inner
        cross-validation folds with the set of parameters that yields the
        best such mean.
    #10: survival predictions for patients in the train set
    #11: survival predictions for patients in the test set

The file that you obtain contains this list saved as a pickled object. You can access the elements of the list by opening the file in python with:

tmp = pickle.load(open(path_to_your_saved_file, 'rb'))

I hope this helps.