The presentation from National Institute of Mental Health shows how to use Pathway Studio data for analysis of fMRI from schizophrenia patients. It finds novel proteins that can be used as biomarkers or drug targets for schizophrenia.

1. Presented at 2017 World Congress of Psychiatric Genetics by:
Dr. Yin Yao , National Institute of Mental Health, USA
Dr. Hongbao Cao, Elsevier, USA
Analysis of Functional Magnetic Resonance
Imaging (fMRI) data from human brain in Pathway
Studio

2. Background and purpose
 Schizophrenia (SCZ) is one of the most chronically
disabling psychiatric illnesses with a global median lifetime
morbid risk of 7.2/1000 persons. Both genetic and brain region
pathways were identified as causal regulating tunnels towards
the pathogenesis of SCZ.
 Integrate Functional Magnetic Resonance Imaging (fMRI)
data set and Brain-Gene ResNet (BGR) data could reveal the
functional pathways through which genes regulate the brain
functions associated with SCZ phenotypes.

3. fMRI data
 functional Magnetic Resonance Imaging (fMRI) data set
 Data were collected with 1.5 T GE MRI scanner;
 Participants were from Kunming, China;
 All 32 patients were treatment resistant;
 All 31 healthy siblings had no schizophrenia history or related symptoms.
 Imaging Data Preprocessing
 SPM8 and REST [1] toolbox: realigned, re-sliced, normalized, band-pass filtered
(0.01~0.08 Hz) and smoothed (3-D Gaussian kernel with 6mm FWHM).
 Whole brain mask was acquired using WFU_PickAtlas toolbox, containing the 116 AAL
brain regions [2].

4. Brain-Gene Resnet (BGR) data in Pathway Studio
 Pathway Studio ResNet ® Mammalian database
http://pathwaystudio.gousinfo.com/ResNetDatabase.html
 Real-time updated network databases, including curated signaling,
cellular processes and metabolic pathways, ontologies and
annotations, molecular interactions and functional relationships.
 Have been widely used to study modeled relationships between
proteins, genes, complexes, cells, tissues and diseases
http://pathwaystudio.gousinfo.com/Mendeley.html
 The largest literature based network database among known
competitors in the field [3].

5. Analysis of fMRI data
 Automated Anatomical Labeling (AAL) based whole
brain connectivity analysis to discover inter-/intra-
brain region dysconnectivity in case of SCZ
Integrate with SCZ ResNet data
 SCZ-gene and SCZ-brain region ResNet relation data
analysis for the nominated SCZ brain regions to
identify potential Genetic-Brain Pathways
Data analysis

6. AAL based full brain CN analysis

7. CR = 81.6% case vs. control with 2 connectivity features within Vermis and
Cerebelum); CR= 74.6% for case vs. sibling with 5 features linked to Middle
occipital gyrus, Fusiform gyrus, Superior parietal lobule, Gyrus rectus and Inferior
temporal gyrus. Color coding represents number of significant connectivity
features associated with each brain region.
fMRI data revealed 7 SCZ brain regions

8. Support from ResNet database
ResNet Database provided literature support of 46
references for 6/7 brain regions for SCZ
(Supplementary Table S1)

9. 4 out of 6 SCZ brain regions were related to 75 SCZ related genes:
Inferior temporal gyrus, Fusiform gyrus, Cerebellar vermis and
Gyrus rectus. (Supplementary Table S2).
Genetic linkage

10. Example: BDNF-Brain-SCZ Pathway
 --Continued

11. Example: BDNF-Brain-SCZ Pathway
BDNF->Fusiform gyrus-> SCZ
 Distortion of the balance among the 3 BDNF isoforms (pro-BDNF,
truncated BDNF and mature BDNF) could lead to changes in
connectivity and synaptic plasticity and, hence, behavior of
Fusiform gyrus [4].
 Deficits in the fusiform gyrus is suggested as a trait pathology in
SCZ patients [5,6].
BDNF-> Cerebellar vermis-> SCZ
 Decreased cerebellar BDNF mRNA and protein level could lead to
motoric impairment and Purkinje cell loss that damage cerebellar
vermis [7].
 Cerebellar vermis is nominated a potential therapeutic target for the
treatment of SCZ [8, 9].

12. Three novel genes for SCZ
 Results suggested 3 novel genes as potential SCZ genes,
including PTEN, FGF8, CD38, supported by 36 studies
(Supplementary Table S3).

13. Three novel genes for SCZ
CD38->Fusiform gyrus->SCZ
CD38 genotype is a genetic factor influencing the function of
Fusiform gyrus[10].
PTEN->Cerebellar vermis->SCZ
Deficiency of PTEN could lead to the loss of neurons and tau
hyperphosphorylation in cerebellar vermis [11].
FGF8 ->Cerebellar vermis->SCZ
Mutations in the FGF8 signaling pathway preferentially affect
the growth of the cerebellar vermis [12].

14. Conclusion
Neuroimaging study alone could gain direct knowledge between
brain regions and the mental health disorders like SCZ
Integrating fMRI data and ResNet data could:
 Generate support for the results from other studies;
 Provide further insights for the understanding of the genetic
mechanism for SCZ;
 Identify novel genes the potentially link to SCZ.
.

15. Future work
 Brain region pathway based fMRI studies (e.g.,
dopaminergic pathways of schizophrenia)
 Genetic pathway – separate brain region
integration
 Genetic pathway – brain region pathway
integration

16. References
 [1] Song, X., et al. PLoS.One. 2011; 6(9): 1-12.
 [2] Tzourio-Mazoyer N, et al., Neuroimage. 2002; 15(1):273-89.
 [3] Lorenzi PI et al. Autophagy. 2014; 10(7):1316-26.
 [4] Kim DW, et al. Schizophr Res. 2013;151(1-3):165-74.
 [5] Walther S, et al. Psychiatry Res. 2009;172(3):184-91.
 [6]Choudhary M, et al. Schizophr Res. 2015 ;162(1-3):103-7.
 [7] Firozan B, et al. Eur J Pharmacol. 2014;732:1-11.
 [8] Villanueva R. Psychiatry Res. 2012;198(3):527-32.
 [9] Garg S, et al. Psychiatry Res. 2016;243:413-20.
 [10] Sauer C, et al. Neuropsychopharmacology. 2012;37(6):1474-82.
 [11] Nayeem N, et al. Mol Cell Neurosci. 2007 Mar;34(3):400-8.
 [12] Wen J, et al. PLoS One. 2013;8(5):e64451.