how to proceed in order to find gene regulation
0
0
Entering edit mode
4.7 years ago
Learner ▴ 250

I have two questions

1- I have a data which is consists of gene expression values 2- I have data which consists of alternative exon splicing

Both data were driven from RNAseq

• Now I am trying to find up/down regulated. I read and I found that one uses Dseq2 normally for RNA-seq analysis. However, the gene expressions are continues values, so do I still need to use Dseq2 or I can use something else?

• I want to know if there will be any problem if I use these values and not taking the raw files (FASTQ) and do all the processes from the beginning?

• which one do you often use to find up and down-regulated and what about the alternative exon splicing? when and how you use them? Please refer to any packages in Python, R, Ruby or Matlab

The gene data look like below

Hybridization REF   1492    1493    1494    1495
Composite Element REF   Signal  Signal  Signal  Signal
C9orf152    3.682158568 3.565245168 3.595608862 3.460994001
RPS11   10.54406969 9.665184958 10.31207154 10.02092773
CREB3L1 6.011743719 6.657238952 5.544938028 6.320113845
PNMA1   8.348027325 8.987468303 9.126783232 9.22168771
MMP2    9.12708295  7.42855084  7.969844335 8.12313501
ERCC5   7.657176246 6.667646814 7.594130938 7.19913373
ZHX3    6.772580276 7.205757758 6.94362774  7.002630483
GPR98   7.387329634 7.477911398 8.937285179 6.020264981
RXFP3   6.771765799 6.965310911 6.559112726 7.181237958
KLHL13  8.419364494 7.269603554 8.823846993 7.864721556
PRO0478 5.15828347  4.964895006 5.594813665 5.469116496
KRTAP10-8   7.201430877 7.03319209  7.126688405 7.216267173
PDCL3   8.938416544 7.676687124 8.106310693 8.001851409
DECR1   8.387471898 8.015597617 8.46314666  8.167643856


and the FIRMA data

Hybridization REF   186 187 188 189 190
Composite Element REF   Score   Score   Score   Score   Score
2315165 -0.21675184 -0.137875837    0.336090128 -0.059906176    -0.244223761
2315391 0.828946771 -0.038886314    0.475539128 0.713607501 -1.215385222
2315396 -0.185979612    -0.172876554    -0.056669873    0.148956577 -0.267353605
2315402 -0.447218229    -0.17860086 -0.21201138 -1.456213417    -0.252945251
2315406 -0.318185645    0.051380891 -0.402628355    -0.366141589    -0.354214602
2315407 -0.344389795    -0.352783141    -0.019746391    1.169567845 0.803986915
2315408 0.290031788 0.705128118 0.757406084 1.727813497 0.640102474
2315410 -1.71043036 0.296647957 0.064062993 -0.001919227    0.290725647
2315425 0.002904956 0.119317475 -0.059390407    0.026593851 -0.177828414
2315433 -1.772023656    0.550938263 -1.066207158    -0.240335476    -1.138320785
2315434 0.269945773 -1.60587177 -2.409227487    1.434148935 0.531043242
2315436 0.127620176 -0.436054086    -0.567006553    0.173804296 0.082854609
2315440 -1.292832369    -0.095422613    -0.435248208    -0.304021258    0.948874618
2315443 0.6826615   -0.527258101    0.31181316  -1.037711509    -0.652166044


Thanks

genomics RNA-Seq • 743 views
2
Entering edit mode

It's not clear what format your data is in - we don't know what your tables look like, and thus it will be difficult to help you. Posting an example record or entry will help us determine what you can do.

DEseq2 is indeed very standard - it uses counts from something like htseq, kallisto, or salmon. They are used to do quasi-mapping and transcript quantification directly from the FASTQ files, which can then be fed into DEseq2 for differential expression analysis.

I don't know what you mean by your gene expression data is continuous, so I can't really answer if DEseq is an appropriate tool for your data or not. Kallisto/Salmon will use whatever mRNA transcripts you feed them for quantitation, so they can handle different transcript isoforms. I don't know what you mean by alternative exon splicing data though. Again, an example would be helpful.

0
Entering edit mode

@jared.andrews07 look above, you can see the example data

2
Entering edit mode

If the data is generated from RNAseq as you mentioned it was processed in some way, which you don't mention and probably don't know. Processing is fine but you should be well aware of what was done and what you should and can further do with the data. For instance, is the data normalized? Is it log scale? I think you should meet with whoever gave you the data and understand what it went through before you'll be able to draw conclusions from.