Pathway analysis

The goal of this post is to utilize the Ingenuity pathway analysis (IPA) to evaluate our gene of interest ACADM.  In previous posts I have described this gene locus and the disease associated with it.  The IPA will be used to search for drugs that may interact with the product of the gene locus and review pathways associated with the locus and its products.

Using the "Genes and Chemicals" search tool and the identifier "ACADM" the entry of our gene of interest was located.  Unfortunately there were no drugs associated with the gene as seen below in the far right cell:

 Additionally linking out to the gene information page also showed the missing drugs table.  This is not surprising given the mechanism of action of the gene and it disease Medium-chain acyl-coenzyme A dehydrogenase deficiency or MCADD.  The disorder manifests itself when periods of fasting occur and the body attempts to utilize fatty acids as a source of energy.  The failure to produce the ACADM protein results in a failure to generate energy and the metabolism starts to fall apart.  This can be prevented through the use of glucose and simple carbohydrates.  As diet has been show to be enough to manage the disease the would be little incentive for a drug targeting this site to be developed.


To begin the analysis of the pathways ACADM was added to a new pathways construct, and using the "build" panel, and selecting the "grow" tools, molecules were limited to on those that have direct interactions, found in humans, and had the molecule types [biologic drug, chemical(8)] excluded.  All molecules both upstream and downstream were allowed, the limit was left at 10 for the first pass of the analysis, but this proved not to be a limiting factor as only two molecules were returned, PPARGC1A and ESRRA, (no trimming was required) as seen here on the left using the "auto-layout".  On the right we see the sub-cellular layout indicating that both molecules act from the nucleus into the cytoplasm on ACADM:

Both of the observed molecules were of the [E] type or expression, meaning that activity by these two protiens act to increase the RNA expression, and the blocking of the activity of these proteins reduced the amount of RNA seen from this locus.  No activation, no inhibition, and no Protein-Protein interaction was seen.
Switching to the "Overlay" tab and selecting Canonical Pathways showed a number of pathways recognized, as seen below:

However only 2 were from ACADM, the Fatty Acid B-oxidation I, and Leucine Degradation I.  As failure of the Fatty Acid B-oxidation I is what actually causes the disease MCADD it was chosen for further analysis.  The view of the pathway can bee seen here:

With ACADM as the purple triangle.  Both of the pathways passing through ACADM are of the type RE, meaning they are enzymatic reaction, part of the break down of fatty acids into energy.  The report for this pathway it is indicated that "Although enzymes of the pathway handle both short and long chain fatty acids, it is the long chain compounds that induce the enzymes of the pathway. Each turn of the cycle removes two carbon atoms until only two or three remain. When even-numbered fatty acids are broken down, a two-carbon compound remains, acetylCoA. When odd number fatty acids are broken down, a three-carbon residue results, propionylCoA. link".  As indicated with the gene itself, no drugs are indicated that interact with this pathway. 

Genome Analysis Part:2

The disease our patient will present with the genome testing, with the disease ACYL-CoA DEHYDROGENASE, MEDIUM-CHAIN, DEFICIENCY OF; ACADMD and with the variant dbSNP:rs77931234 (as discussed in previous posts) in the gene ACADM.

In the NCBI browser the variant of interest can be seen here at position 75761161 in GRCh38:

here a larger view can be seen of the exon the variant is in:

The location for GRCh37 and GRCh38 can be seen here from the dbsnp entries:

The VCF entry for a homozygous mutation would look as follows:

#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT CB00001
1 75761161 rs77931234 A G 25 PASS NS=1;DP=1;5;DB GT:GQ:DP 1/1:52:35

Genome Analysis Part:1

The goal of this post is to view the current state of clinical whole genome sequencing (wgs) and clinical whole exome sequencing (wes) from the point of view of an individual suffering in a disease state and with a diagnosis that does not completely explain their symptoms.   

The patient in this example has been previously diagnosed with a Medium-chain acyl-coenzyme A dehydrogenase deficiency, through phenotype characterization which was confirmed with molecular diagnosis (PCR) at a reference lab to be homozygous for the mutation rs77931234.  The diagnosis was made when he was young, and his disease has been well managed.  Around puberty new symptoms began to appear which are not normally associated with the known disease phenotype.  The patient and family has tried a variety of medications with mixed results often resulting in reappearance of symptoms following six months or so of stability.  A battery of both metabolite and genetic tests have been run with no clear diagnosis, but metabolite values are often outside of normal ranges. The family has been on a classic "diagnostic odyssey" with no clear explanation.  

Fortunately for this family are patients at a regional medical center in a large metropolitan area and have been  referred to the Individualized Medicine Clinic.  After being seen in the clinic, and returning for a follow up the family is presented with 4 options for genome scale testing in an attempt to find possible genetic causes for the additional symptoms:

1 - Participation in a clinical trial offering full exome analysis for the patient and the parents at no cost
2 - Run a full genome for the patient only and try to get costs covered by insurance(4-6 months)
3 - Run a full genome on the patient and pay out of pocket ($5-$10k)
4 - Use a direct to consumer service and perform analysis of the raw results

The family unsure of which option is best, has sent an email describing the situation and asking for advice on what they should do.  

My recommendation will be colored by my experience as a bioinformatician working in a genome center who has had experience with research protocols, clinical genotyping, WGS and WES analysis.  I would strongly recommend the family go with option 1 "clinical trial offering full exome to the family", for the following reasons.

While we are quickly approaching a future of genome enabled medical care, with deep understanding of how the genome works and changes affect health, we have not reached that point.  This is the primary problem with options 2 and 3.  These options both have the additional problem of "who owns the data", if the patient pays for it and how would a sequencing center deliver this information for interpretation?  If the insurance pays for it, do they have a right to dig through the data and adjust your rates based on the findings? Sending a hard drive to a patient (or a clinician) who is unfamiliar with the standard file formats or tools used to evaluate data is not a real solution.  The additional complication of a genome returning on the order of 4-5 million variants or more, makes manual interpretation impossible.  Additionally as wgs tools advance and becomes more common a deeper understanding of the genome will emerge, however we currently rely on the data compiled in the previous century which focused heavily on gene, and pathway analysis and how single variants would impact this system, making most of the variants associated with pathogenic conditions contained inside of the gene sequence itself, leaving much of the intergenic sequence regions dark and unexplored, and variation in these regions poorly characterized.  Option 4 (direct to consumer) while possibly cheaper, means you have to find an informatician and possibly contract them to do the analysis, and then follow up with finding a clinician to return some interpretation of the variants found.  The problem is compounded by the variety of vendors, price and quality of data produced.  23andME is actually a very reliable data producer, unfortunately interpretation of results for medical care is hamstrung by the FDA's recent threats against the company.  Those companies that provide wgs or wes as a service require careful evaluation as many cut costs (and corners) by providing a lower threshold of data returned (say 10x coverage instead of the industry accepted 25-40x for wgs) reducing the confidence of calls in complex gene regions.   

Many of these problems are solved by option 1, the clinical exom trial.  First it limits data analysis to gene exons which is where most of our variant knowledge exists meaning that variants found through this tool will have likely been seen and characterized in a publicly available database such as OMIM, dbsnp, or Exome variant server ect. The recommendation of exome trio analysis is also part of the presentation.  Having the parents sequenced in parallel with the child allows for the removal of the variants which were inherited from two unaffected a parents leaving those that are likely the source if the disease is genetic in origin.  While trio analysis also has a long way to go for the detection of hetrerozygous gene drop out, it remains one of the most powerful tools in genetic analysis. As part of a clinical study implies that there will be other sequencing runs done and having the data be part of a well quality controlled pool and workflow is just as important as the trio analysis.

While I would love to recommend that everybody get a wgs analysis done, for a single patient trying to make sense of the complex and rapidly changing world of genetic analysis the current state of the art would suggest exome analysis as the most reasonable method with the greatest clinical application.